r 
 
 REESE LIBRARY 
 
 OF THE 
 
 UNIVERSITY OF CALIFORNIA 
 
 ccession No. 
 
THE PURIFICATION 
 
 OF 
 
 PUBLIC WATER 
 SUPPLIES 
 
 BY 
 
 JOHN W. HILL 
 1 1 
 
 CONSULTING ENGINEER 
 
 MEMBER AMERICAN SOCIETY OF CIVIL ENGINEERS, MEMBER AMERICAN WATER WORKS ASSOCIATION, 
 MEMBER AMERICAN PUBLIC HEALTH ASSOCIATION 
 
 NEW YORK 
 D. VAN NOSTRAND COMPANY 
 
 LONDON 
 E. & F. N. SPON, 125 STRAND 
 
 i 
 
COPYRIGHT, 1898, 
 BY D. VAN NOSTRAND COMPANY. 
 
 TYPOGRAPHY BY C. J. PETERS & SON, BOSTON. 
 
 Plimpton pres 
 
 NORWOOD, MASS. 
 

 PREFACE. 
 
 THIS work is in continuation of a series of lectures and papers 
 on the Quality of Public Water Supplies, which the author has 
 had the honor to read before several scientific societies and uni- 
 versities during the past five years, and is intended to present in 
 a brief way, (i), the fact and causes of pollution of sources of 
 public water supply; (2), the effect of this pollution on the 
 typhoid fever rates of our larger cities ; and (3), to illustrate by 
 a few examples how the typhoid rates have been reduced by the 
 introduction of water from purer natural sources and by filtration 
 of polluted waters. In connection with the subject of water 
 quality, brief reference is made to the water bacteria, and some 
 data are given on the methods of construction and operation of 
 sand filters, together with the cost of filter construction in dif- 
 ferent water-works, and the cost per million gallons of water 
 treated. 
 
 The principal object in bringing out this work at the present 
 time is to impress upon city officials, health officers, and others 
 connected with or interested in works of public water supply, the 
 necessity of a more vigorous attack of the problems of "Water 
 Pollution," and "Purification of Water" intended for drinking and 
 other dietetic uses. 
 
 The statistics of population and typhoid fever death rates 
 (Appendix A) have been obtained from health officers, water- 
 works managers, and official published reports. 
 
 The author has endeavored to acknowledge all sources of 
 information in the body of the book ; but especial thanks are due 
 Dr. Dunbar of the Hamburg Hygienic Institute for several valu- 
 able original papers on water supply in Germany ; Mr. Rud Schro- 
 der, inspector of the Hamburg water-works, for much valuable 
 
iv PREFACE. 
 
 information on the operation of these filters ; and Mr. Thomas 
 W. Boughen of Cincinnati, who by the author's request kindly 
 undertook to collect exact information upon the details of filter 
 construction and operation during a recent tour of Europe. 
 
 In the translation of German, French, and Spanish papers and 
 reports on European water-works and the hygiene of water, he 
 has had kind and valuable assistance from Dr. Philip Hillkowitz, 
 Charles E. Rasinsky, and Adolph G. Wulff, graduates of the Uni- 
 versity of Cincinnati. Valuable assistance has also been rendered 
 by Mr. Paul Hamilton, a graduate of the University of Michigan, 
 in the preparation of tabular matter on recent filter practice, etc. 
 
 All the illustrations, excepting Figs. 10, 11, and 12, were pre- 
 pared especially for this work by his son, Mr. Henry C. Hill, to 
 whom the author is also largely indebted for patient assistance, 
 and many valuable suggestions in connection with the experimen- 
 tal work and collation of authorities for matter which appears in 
 the book. 
 
 Chapter I. was originally read as a lecture before the Cincinnati 
 Section of the American Chemical Society at the Cincinnati Uni- 
 versity, Jan. 15, 1897 ; and Chapter II. was originally read as a lec- 
 ture before the Academy of Medicine, Cincinnati, May 3, 1897. 
 Chapter IX. was originally read as a paper on the "Sterilization of 
 Drinking- Water as a Means of Reducing the Typhoid Fever Rates " 
 
 at Buda-Pest, September, 1894. 
 
 j. w. H. 
 
 CINCINNATI, June, 1897. 
 
CONTENTS. 
 
 CHAPTER I. PAGE 
 
 INTRODUCTION 1 
 
 Water the main constituent of the animal system. Difficulty of obtaining exact proof 
 of water quality. Modes of test admissible in physics not applicable to test of 
 water quality. Changes in character of water source by pollution. General confi- 
 dence in present standards of water quality still to be established. Absolutely 
 pure water not found in nature, nor is it essential for dietetic purposes. In the 
 light of present information cities are not justified in supplying polluted water 
 to consumers. Hygienic laboratory at Hamburg. Management of filters in Hol- 
 land. Sources of water supply to the cities of Manchester and Vienna. Work of 
 Massachusetts State Board of Health in behalf of public water supplies. The fil- 
 ters of London thirty years ago. Zymotic disease and prophylaxis. Transmission 
 of pathogenic organisms through the medium of water supply. Failure to find 
 the typhoid bacillus in polluted water not to be taken as evidence of its non-exist- 
 ence. Pure water cannot create disorders of the animal system. Pure water better 
 than purified water. Filtration cannot render water absolutely pure. Typhoid 
 fever rates of London, Munich, Berlin, Chicago, Pittsburg, Louisville. Habits 
 of people in the United States and Europe. Interest in water supplies of high 
 hygienic quality by the large cities of Europe. 
 
 CHAPTER II. 
 SOURCES OF PUBLIC WATER SUPPLY 14 
 
 Sources of water supply for cities. Hard water from limestone watersheds. Other 
 sources of water pollution than sewage. Professor E. Ray Lankester's opinion on 
 b. typhosus and b. colt communis. Rivers and lakes receiving surface drainage 
 cannot be regarded as uncontamihated water sources. Sources of satisfactory 
 water supply at high elevations of rare occurrence. Rivers constitute the largest 
 general source of water supply for cities. Dilution of sewage-polluted waters can- 
 not be depended upon for the removal of disease germs. Longevity of typhoid 
 bacillus in water. Sewage and garbage should not be disposed of in rivers and 
 lakes which constitute sources of water supply. Dr. G. Sims Woodhead on the 
 avoidance of surface drainage into the Thames. Professor Baumeister on the influ- 
 ence of dilution of sewage. Lakes subject to same sources of water pollution as 
 are rivers. Improvement of water quality by sedimentation. Experiments by 
 Dr. Miquel. Lake Zurich as a source of water supply. Self-purification of sewage- 
 polluted waters cannot be seriously entertained. Dr. Drown's experiments on 
 aeration of polluted waters. Theory of self-purification of rivers and lakes. Im- 
 pounding reservoirs not safe sources unless drainage ground is laid waste. Pollu- 
 
VI CONTENTS. 
 
 PAGE 
 
 tion of dug wells. Ground water may be well purified before it is intercepted by 
 wells. The typhoid fever rates as an index of water quality. Feasibility of prevent- 
 ing direct sewage pollution of water sources. Filtration to be successful must meet 
 all the varying conditions of the unfiltered water. Failures in filtration due to 
 ignorance and carelessness. High rates of filtration not admissible. Deep driven 
 wells not always proper sources of water supply. Natural filtration through mate- 
 rials in the drift not always complete. Dr. Rosenau's investigation of deep well 
 water supplied to San Francisco. B. proteus vulgaris in driven well water. Dis- 
 trust of nearly all natural sources of water supply by foreign investigators. Chol- 
 era a disease not indigenous to this country. Typhoid fever the principal disease 
 to be restrained by pure water or water purification. Objection to water otherwise 
 pure, but high in mineral constituents. Sentiment against polluted water not as 
 strong as it should be. Tests of water quality. Comparison of water supplies of 
 Jersey City and Newark, N. J. Water supply of Lowell and Lawrence, Mass. 
 Dr. Rogers on causes of mountain fever. Mr. Preller on pollution of mountain 
 water by droppings from cattle. Many diseases of domestic animals are also 
 diseases of man. Water the only article of diet of universal use. Other causes 
 than water responsible for some typhoid epidemics, but polluted water is regarded 
 as the cause of high continuous typhoid rates. 
 
 CHAPTER III. 
 BACTERIAL CONTENTS OF VARIOUS WATERS 40 
 
 Causes of variation in numbers of bacteria from same source. Bacteria in cistern 
 water. Bacteria in Ohio River water. Longevity of pathogenic bacteria increased 
 by dilution of sewage. Bacterial contents of deep well water. Determinations by 
 Professor Sedgwick. Bacterial tests of deep well water by author. Bacteria in water 
 from Pasteur filters. Freudenrich's tests of Pasteur filters. Author's tests of 
 water from Pasteur filters. Stone disk and tube filters. Bacteria in water from 
 stone disk and tube filters. Influence of days of growth on numbers of bacteria in a 
 water sample. Spring waters. Bacterial examinations by Dr. Drown. Bacterial 
 examinations of spring water by Professor Sedgwick. Tests of the Ohio River 
 water in Parietti solution. Rainwater. Bacterial tests by Dr. Miquel. Bacterial 
 tests by author. Chemical contents of rainwater by Dr. Drown. Bacterial con- 
 tents of distilled water. Bacterial contents of a water suspected of sewage pol- 
 lution. Bacterial contents of water from small Anderson Iron Purifier and sand 
 filter. Bacteria in artificial ice. Bacteria in the air. Influence of sunlight on 
 bacteria in water. 
 
 CHAPTER IV. 
 THE TYPHOID BACILLUS AND TYPHOID FEVER 56 
 
 Properties of b. typhosus. Non-pathogenic organisms which resemble in some re- 
 spects b. typhosus. Comparison of b. typhosus. b. coll communis, and b. lactis 
 aerogenes. Experiments on b. typhosus and b. coli communis in sterilized milk. 
 B. coli communis and b. typhosus in polluted water. Dr. Alessi's experiments 
 with putrid gases on rats, guinea pigs and rabbits. Direct proof of the presence 
 of b. typhosus in a water supply not essential. Sanarelli's papers on the etiol- 
 ogy of typhoid fever. Typhotoxin. Dr. Jordan on the identification of the 
 typhoid bacillus. Mortality from typhoid fever. Seasonal distribution of ty- 
 phoid fever. Reduction of typhoid in Munich. Dr. Reincke on typhoid fever 
 in Hamburg and Altona. The large cities as typhoid fever centers. 
 
CONTENTS. Vll 
 
 CHAPTER V 
 
 PAGE 
 
 CLASSIFICATION OF CITIES BY TYPHOID FEVER STATISTICS ... 70 
 
 Final test of water supplies is the influence on the public health. Cities of the first 
 class. Cities of the second class. Cities of the third class. Cities of the fourth 
 class. Cities of the fifth class. Cities of the sixth class. Cities of the sev- 
 enth class. No city of the United States in the first class. A reduced death rate 
 from typhoid fever easy of attainment if desired by municipal authorities. Water 
 supplies of Rotterdam, Amsterdam, and The Hague. Water supplies of Vienna, 
 Munich, Dresden, Berlin, London, Edinburgh, New York, Brooklyn, Hamburg. 
 On the use of beer and wine by people of Europe and the United States. 
 
 CHAPTER VI. 
 PURE AND PURIFIED WATERS 81 
 
 Superiority of pure and purified waters. Sources of pure water very rare. Definition 
 of "pure" and ''purified" waters. Degree of water "purity." Protection of 
 watersheds. Consumption and waste of water in American cities. Sources of 
 ground water supply. Double system of public water supply. Advantage of water 
 supply from sources of natural purity. Filtration of impounded and surface waters 
 for public supply. Pollution of deep well water. Objections to dual water sup- 
 plies. Influence of sterilized water on the human system. Poisonous mineral 
 matters in ground and surface waters. Storage of ground and surface waters. In- 
 fluence of sunlight on the growth of algae. Comparison of water supplies of small 
 and large communities. Sources of water supply of Vienna, Munich, Dresden, etc. 
 Use of water from mechanical filters and driven wells. Revenue the dominating 
 factor in the location of cities. Feasibility of procuring satisfactory water from 
 natural sources. 
 
 CHAPTER VII. 
 CITATIONS ON TYPHOID FEVER EPIDEMICS 91 
 
 Value of practical illustrations of the relation of water quality and typhoid fever. 
 Epidemic at Lausen. Epidemic at Caterham. Epidemic at Plymouth, Penn. Epi- 
 demic at Zurich. Epidemic at Spring Water, N.Y. Epidemics at Lowell and 
 Lawrence, Mass. Epidemic at Sault Ste. Marie, Mich. Analysis of Sault Ste. 
 Marie water. Epidemic at St. Louis, Mo. Epidemic at Elmira, N.Y. Relation 
 of San Francisco water and typhoid fever. Typhoid rates of Denver, Col. Epi- 
 demic at Middletown, Conn. Epidemic at Stamford, Conn. Epidemic at Eliza- 
 beth, N.J. Typhoid fever at Evansville, Ind. 
 
 CHAPTER VIII. 
 SEDIMENTATION OF POLLUTED WATERS 110 
 
 Experimental information rather meager. Miquel's experiments with the Seine 
 water. Limited subsidence can produce no marked change in quality. Effective 
 sedimentation in large deep reservoirs. Reduction of bacteria in Ohio River water 
 by subsidence. Reduction of bacteria in water from Lake Linthrathen by subsi- 
 dence. Reduction of bacteria in London water by subsidence. Experiments with 
 alum and slaked lime on Ohio River water. Effect of lime process on Colne Valley 
 water. Experiments by Professor Lankester with mud and clay in Oxford water. 
 Experiments by Professor Lankester with alum and lime in Oxford water. Experi- 
 
Vlll CONTENTS. 
 
 ments by Mr. Flad on sedimentation of Ohio River water. Rate of subsidence of 
 suspended matter. Experiments by Mr. Dibdin with lime in the New River water. 
 Influence of lime for reduction of hardness on the bacterial contents of water. Cost 
 of applying the lime process to the water supply of London. 
 
 CHAPTER IX. 
 STERILIZATION OF DRINKING-WATER . 120 
 
 Purification of drinking-water should be conducted by the municipal corporation or 
 water company. Use of boiled water in cases of doubt. Filtered and boiled 
 water. Investigation of the method and cost of sterilizing the dietetic water for a 
 community. The Yaryan process of water sterilization. Distribution of sterilized 
 water through separate system of small mains. Separate services for sterilized and 
 unsterilized water. Construction of public improvements in the United States 
 influenced by political considerations. Use of sterilized water by employees at 
 the Columbian Exposition. Objections which have been raised to a double supply. 
 
 CHAPTER X. 
 FILTRATION OF WATER SUPPLIES 131 
 
 Works for treatment of water for city use. Domestic filters not a safeguard against 
 contaminated water. Continuous sand filtration in Europe. Filtration cannot 
 produce absolutely pure water. Theory of filtration. Biologic action of a sand 
 filter. Scraping and aeration of sand-beds. Action of bacteria on organic matter 
 in water. Penetration of sand-bed by bacteria. Experiments by Piefke with sand 
 filters at Berlin. Head on Lake Miiggel filters. Refilling a scraped filter. Inter- 
 mittent sand filters. Theory of action. Influence of the nitrifying bacteria on 
 organic matter in water. Lawrence, Mass., intermittent filter. Natural filtration 
 in the materials of the drift. Comparison of pure spring water with purified river 
 water. Artificial filtration more reliable than natural filtration. Dr. Drown's 
 opinion of natural filtration. Sedimentation and filtration in Europe. Time al- 
 lowed for sedimentation in different cities. Mr. Binnie on filtration of water from 
 Welsh sources. Filters of Altona, Germany. Mr. Kiimmel on maximum rate of 
 filtration. Rate of filtration at the Lawrence, Mass., experiment station. Rate 
 of filtration as affecting the quality of Zurich water. Bacteria in London waters. 
 Operation of the London filters (1896). Bacteria in Hamburg water. Bacteria in 
 Lawrence water. Double filtration. Filtration should be measured by practical 
 results. Influence of storage on waters. Covered and open reservoirs for filtered 
 water. 
 
 CHAPTER XL 
 TYPES OF SAND FILTERS 158 
 
 Classification of sand filters. Filters with vertical walls of masonry. Filters with 
 sloped walls of earth. Regulating devices. Advantage of sloped walls for open 
 filters in cold climates. Arrangement of drains and filtering materials. Capacity 
 of a filter. Scraping the surface of a sand-bed. Rate of delivery. Grading of 
 filtering materials. Effective size of sand grain. Uniformity coefficient. Sterili- 
 zation of filter sand. Lawrence, Mass., intermittent filter. Rate of filtration. 
 Periods of work and intermission. Expense of operation for Lawrence filter for 
 1895. Rate of filtration for 1895. Bacterial results for 1895. Cost of Lawrence 
 filter. Mr. Shedd on sand filtration. Providence, R.I., experiments with sand 
 filters. Bacterial efficiency of plain sand filter with alum. Lowell, Mass., filter. 
 Hudson, N.Y., filters. Poughkeepsie, N.Y., filters. Filter galleries. Filters of 
 
CONTENTS. ix 
 
 Rotterdam. Filters of The Hague. Filters of Amsterdam. Bilters of Paris sub- 
 urbs. Anderson Iron Process. Filters of Zurich. Extent of London filters. 
 
 CHAPTER XII. 
 MECHANICAL FILTERS . . >. . 184 
 
 Operation. Comparison of rates with London plain sand niters. Types of mechani- 
 cal filters. Experiments at Providence, R.I. Morison mechanical filter. Worst re- 
 sults a measure of efficiency of filtration. Reduction of color by mechanical filters. 
 Results with and without alum. Cost of filters. Cost of operation compared with 
 plain sand filters. Use of mechanical filters in American cities. Somerville and 
 Raritan, N.J., filters. Long Branch, N.J., filters. Lorain. Ohio, filters. Mechan- 
 ical filtration, Albany, N.Y. Mechanical filtration, Philadelphia, Penn. Elmira, 
 N.Y., filters. The use of alum for filtration. Influence of alum on the human 
 system. Consumption of alum in filtration. Drs. Thomas and Marshall on alum 
 filtration for Philadelphia. Mechanical filtration works in cities of America. 
 Reduction of iron by mechanical filters, Asbury Park, N.J. Dr. Dunbar on reduc- 
 tion of iron in German ground waters. Reduction of iron in ground water, Read- 
 ing, Mass. 
 
 CHAPTER XIII. 
 
 HAMBURG SETTLING-BASINS AND FILTERS 208 
 
 History. Settling-basins. Conduit from settling-basins to filters. Influent cham- 
 bers. Description of filters. Collecting-drains. Arrangement of filtering materials. 
 Washing of filtering materials. Quantity of filtering materials. Effluent chambers 
 and regulating weir. Measuring the effluent. Determining the size of filtration 
 works. Operation of the filters. Cleaning a filter. Refilling a filter. Clear- 
 water basin. Period of operation of filters. Sand-washing machinery. Schroder 
 sand washers. Capacity of sand washers. Ice on filters. Mode of cleaning ice- 
 covered filters. Mager sand-scraper. Comparison of old and new methods of clean- 
 ing ice-covered filters. Method of scraping sand during winter 1896-1897. Cause 
 of typhoid epidemic, autumn of 1897. 
 
 CHAPTER XIV. 
 
 THE FILTERS OF THE BERLIN WATER-WORKS 230 
 
 Works at Stralau and description of filters. Works at Lake Miiggel and description 
 of filters. Regulation of filters. Cost of covered filters. Clear-water reservoir. 
 Operation of filters. Sand-washing machine. Bacterial efficiency of filters. 
 
 CHAPTER XV. 
 THE FISCHER FILTER AND ANDERSON PURIFIER 238 
 
 Fischer filter at Worms, Germany. Cost of filter compared with plain sand filters. 
 Anderson Revolving Iron Purifier. Experimental results obtained with the Ander- 
 son Purifier at Paris. Dr. Dupre on Anderson Purifier at Worcester, Eng. Cost 
 of purifying water by the Anderson process. Cost of installation of Anderson 
 Purifiers. 
 
 CHAPTER XVI. 
 
 FILTERS PROPOSED FOR CINCINNATI . 246 
 
 Condition of Ohio River water. Experiments on Ohio River water. Subsiding-reser- 
 voirs. Description of filters. Arrangement of filtering materials. Regulating- 
 chambers. Clear well. Open and closed filters. Cost of original and amended 
 plans of filters. Cost of Berlin and Hamburg filters. 
 
X CONTENTS. 
 
 CHAPTER XVII. 
 
 PAGE 
 
 COST OF FILTERS AND FILTRATION 255 
 
 Conditions effecting cost of works of filtration. Estimated cost of filters for Philadel- 
 phia. Estimated cost of filters for Providence. Mr. Hazen's estimate of cost of 
 covered filters. Estimated cost of filters for Cincinnati. Open filters. Covered 
 filters. Estimated cost of filters for Albany. Capacity of clear well. Estimate on 
 a system of ten filters and clear well. Cost of Berlin and Hamburg filters. Cost 
 of Zurich filters. Cost of Lawrence filter. Rates of filtration per acre per day. 
 Rates proposed as standards for plain sand filters. Duration of service of filters. 
 Scraping of sand-bed. Rotation of sand-bed. Cost of scraping the London filters. 
 Washing sand for filters. Cost of filtration. Estimate of cost by Mr. Hazen. 
 Estimate of cost for Philadelphia. Estimate of cost for Albany. Cost of filtra- 
 tion at Zurich. Estimate of cost for Cincinnati. Cost of filtration at London. 
 Cost of filtration at Poughkeepsie, N.Y. Cost of operating filters in Germany. 
 Estimate of cost of filtered water per capita per annum. 
 
 APPENDIX A. 
 TYPHOID FEVER STATISTICS FROM LARGE CITIES OF THE WORLD, 268 
 
 APPENDIX B. 
 THE BACTERIA 272 
 
 Chemical composition. Mycoprotein. Products of bacterial action on organic mat- 
 ter. Saprophytes and parasites. Liquefiers and non-liquefiers. Aerobians and 
 anaerobians. Forms of bacteria. Cocci. Bacilli. Spirilla. Motility of certain 
 species. The flagella. Putrefactive and pathogenic bacteria. Chromogenic spe- 
 cies. Ptomains. Toxins. Measurement of the bacteria. Bacteria in air, soil 
 and water. Putrefactive organisms from sewage sources. Spore-bearing bacteria. 
 Number of species found in water. Pathogenic bacteria found in water. No 
 proof of the diphtheria bacillus being transmitted by water. Staining-properties 
 and differentiation. Effect of high temperatures on water bacteria. Bacteria 
 resembling in some respects b. typhosus found in water. Bacteria smaller than 
 b. typhosus found in water. Large bacteria found in water. Spore-bearing bacteria 
 found in water. 
 
 APPENDIX C. 
 
 THE LEGAL LIABILITY OF CITIES AND WATER COMPANIES FOR DAM- 
 AGES BY SEWAGE POLLUTED WATER 287 
 
LIST OF ILLUSTRATIONS. 
 
 FIG. PAGE 
 
 1. TYPHOID FEVER RATES NEWARK AND JERSEY CITY, N.J. BLACK, 
 
 NEWARK ; SHADED, JERSEY CITY 33 
 
 2. TYPHOID FEVER AND RAIN-FALL FROM JANUARY, 1882, TO DECEMBER, 
 
 1895. (SAN FRANCISCO, CAL.) 105 
 
 4. I YARYAN APPARATUS FOR STERILIZING WATER 127 
 
 5.J 
 
 6. DIAGRAM SHOWING ACCUMULATION OF BACTERIA IN SAND-BED . . 135 
 
 7. DIAGRAM SHOWING OPERATION OF LONDON FILTERS 151 
 
 8. DETAILS OF PROPOSED FILTER FOR ST. Louis WATER- WORKS. (KIRK- 
 
 WOOD) 155 
 
 9. LONGITUDINAL SECTION OF FILTER. (LAWRENCE, MASS.) .... 167 
 
 10. EXPERIMENTAL FILTER. (PROVIDENCE, R.I.) 170 
 
 11. JEWELL GRAVITY FILTER 185 
 
 12. MORISON EXPERIMENTAL FILTER. (PROVIDENCE, R.I.) 186 
 
 13. INFLUENT CHAMBER SHOWING REGULATOR. (HAMBURG, GER.) . . 209 
 
 14. SECTION OF MAIN DRAIN AND FILTERING MATERIALS. (HAMBURG, 
 
 GER.) 211 
 
 15. PLAN OF FILTER. (HAMBURG, GER.) 212 
 
 16. EFFLUENT CHAMBER SHOWING WEIR. (HAMBURG, .GER.) .... 215 
 
 17. DIAGRAM SHOWING OPERATION OF FILTER No. 12. (HAMBURG, GER.), 216 
 
 18. DIAGRAM SHOWING OPERATION OF FILTER No. 16. (HAMBURG, GER.), 217 
 
 19. SAND- WASHING PLANT, HOPPER No. 1. (HAMBURG, GER.) . . . 221 
 
 20. SAND- WASHING PLANT, HOPPER No. 2. (HAMBURG, GER.) . . 221 
 
 oo'l 
 
 03 f SAND-WASHING PLANT. (HAMBURG, GER.) 222 
 
 24J 
 
 25. DIAGRAM SHOWING ICE ON FILTERS, WINTER OF 1896-1897. (HAM- 
 
 BURG, GER.) 224 
 
 26. DEVICE FOR SCRAPING ICE-COVERED FILTERS. (HAMBURG, GER.) . 225 
 
 27. PLAN OF FILTERS AT LAKE MUGGEL. (BERLIN, GER.) 231 
 
 28. PLAN OF REGULATING-CHAMBER. (BERLIN, GER.) 233 
 
 29. SAND-WASHING MACHINE. (BERLIN, GER.) 236 
 
 30. FISCHER FILTER. (WORMS, GER.) 239 
 
 nnt i ANDERSON REVOLVING IRON PURIFIER 242 
 
 oOt>. l 
 
 31. ^ 
 
 32rt. IPROPOSED PLAN FOR OPEN FILTERS. (CINCINNATI, O.) . . . . 248 
 32^. J 
 
 xi 
 
THE PURIFICATION OF WATER, 
 
 CHAPTER I. 
 
 INTRODUCTION. 
 
 WATER is an essential of human existence. According to 
 Landois,* 58.5 per cent of the body weight is water, and nearly 70 
 per cent of the blood corpuscles is water ; of the serum of the 
 blood, 90 per cent is water, f No other article of "diet enters so 
 completely into the construction and support of the animal system. 
 A very early writer held that the blood of an animal was life, $ and 
 the use of it as an article of food was interdicted by holy law. 
 Water being the main constituent of the blood, it may also be 
 regarded as the principal element of animal life ; and all diligence 
 should be exercised in procuring for dietetic purposes a water 
 which, while readily assimilable by the system, shall not be the 
 cause of disease. 
 
 It is a curious fact, borne out by the many costly and pains- 
 taking investigations into the resources of cities for public water 
 supply, that, after all, we are not thoroughly informed upon the 
 question of water quality. This is not in disparagement of the 
 labors of the many able investigators along this line of scientific 
 research ; but while in nearly every other branch of physics satis- 
 factory proof of certain qualities of matter can be had, the absolute 
 proof of the hygienic quality of water supplies is still beyond the 
 reach of our most modern methods of research. If one is disposed 
 to question this statement, his careful attention is invited to the 
 
 * Landois' Human Physiology. 
 
 t Transactions American Society of Civil Engineer 'S, vol. xxxii., p. 151. 
 
 J Genesis, chap, ix., v. 4. 
 
 1 
 
2 THE PURIFICATION OF WATER. 
 
 views of some of the ablest water analysts of England, as shown 
 by the exhaustive investigation of the Royal Commission on 
 Metropolitan Water Supply.* 
 
 If one is in doubt as to the strength of a bar of steel, he can 
 easily resolve his doubts by putting a specimen into a testing- 
 machine and breaking it. The results will satisfy him upon all 
 the physical properties of the metal. If he desires to pursue the 
 inquiry further, he can obtain very satisfactory evidence of the 
 composition of 'this steel, and reasoning a priori, can make as 
 many bars substantially like his specimen as may be desired. 
 
 Examinations of water samples, however, are not so satisfac- 
 torily conducted. The results obtained at one time are not often 
 verified by subsequent tests. Changes in the chemical and bio- 
 logical condition of the water are constantly going on ; and it is 
 not unlikely that a public water supply might comply with all the 
 recognized standards of potability at one time, and be subject to 
 just condemnation at another. And right here lies the danger 
 to communities which depend upon water from a common source. 
 
 The experience at Plymouth, Penn. (1885), shows how a hith- 
 erto satisfactory source of water supply may become an agent of 
 destruction, with no preliminary indications of the time or nature 
 of the changes which were taking place in the previously pure 
 water of this little mountain reservoir. 
 
 The terms " pure " and " impure " with reference to water are 
 used advisedly. If the water is safe for drinking and dietetic pur- 
 poses, it is "pure" although such water, if from natural sources, 
 would not be found chemically and bacterially " pure ; " while an 
 "impure" water is one that is the cause of disease, even though 
 the chemist and bacteriologist might not be able to decide upon 
 the evidence or nature of the impurity. Impurities may come 
 into water from the atmosphere, from surface drainage, and from 
 sewage. But the impurities which are feared in water are the 
 pathogenic and putrefactive bacteria, and the ptomains. 
 
 The pathogenic bacteria are those specifically concerned in 
 disease, and held to be a part of its etiology. The putrefactive 
 bacteria found in all water rich in organic matter, especially from 
 
 * London, Eyre & Spottiswoode, 1893. 
 
JNTR OD UC TION. 3 
 
 sewage sources, may produce disorders of the digestive tract, 
 although not held to be the specific agents of disease. The 
 ptomains have never within the author's knowledge been found in 
 water, although it is reasonable to suppose that such may come 
 into water from putrefying organic matter lying upon the fore- 
 shores of rivers and lakes ; but the dilution of these will be very 
 great in all ordinary instances of rivers and lakes constituting 
 sources of water supply. 
 
 The biologist and bacteriologist deal in matter found only in 
 suspension in water ; and dangerous substances may exist in solu- 
 tion, and their methods of search would not disclose the fact. If 
 ptomains ever occur in a water supply they will be in solution, and 
 the ordinary chemical water analysis will not reveal them. Indeed, 
 the proof by chemical means of a ptomain or toxic substance in 
 water will be found upon investigation to be very difficult, if not 
 altogether impossible, by any known process.* It is a disputed 
 question whether Brieger and his co-laborers have really precipi- 
 tated the toxic substances of bouillon cultures of the pathogenic 
 bacteria,f and altogether it may be held that the absolute proof of 
 water quality is still an unattainable result. 
 
 This fact, however, should not diminish the perseverance of 
 the workers in the field of water analysis and purification, but 
 rather serve as a stimulus to stronger and higher efforts in behalf 
 of the thousands who annually perish from water-borne diseases. 
 
 That certain waters, when judged by our present standards, are 
 held to be safe for drinking and other dietetic purposes, while 
 other waters, judged by the same standards, are held to be unfitted 
 for such uses, we all know ; but general confidence in these same 
 standards is still to be established. 
 
 Absolutely pure water is not found in nature. All water, from 
 whatever source, even freshly fallen rain-water, contains some evi- 
 dences of contamination ; but according to our standards, water 
 fit to drink is often found in natural sources. It is not essential 
 that water for drinking and dietetic purposes be chemically and 
 
 * Royal Commission on Metropolitan Water Supply, London, 1893. Professor E. Ray 
 Lankester, Appendices to Evidence, p. 452. 
 
 t Annales de rinstitut Pasteur, Sanarelli, April, 1894. 
 
4 THE PURIFICATION OF WATER. 
 
 bacterially pure ; but it is essential that it contains no pathogenic 
 organisms, and shall be free from ptomains due to the action of 
 bacteria upon decaying organic matter. Whether the latter have 
 really occurred in drinking-water is not certainly known, but some 
 investigators at the present time seem to suspect the possibility 
 of it. 
 
 The pumping of water for domestic uses from a source known 
 to be polluted by sewage or otherwise should be severely con- 
 demned. The delivery of water containing the elements of fatal 
 disease to a confiding and helpless community should be ranked 
 with the sale of intoxicating liquors to minors and confirmed in- 
 ebriates. An attempt to kill people by the systematic distribution 
 of a poison would be met by the apprehension and punishment of 
 the offender ; while the delivery of water for drinking and other 
 dietetic uses, as fatal to some as a dose of strychnine, is going on 
 in nearly every large city of the land. Shall we shut our eyes to 
 the fact that polluted water is dangerous to health, or shall we 
 recognize the evil, and address ourselves to its remedy ? 
 
 Every city which continues to supply a tainted water without 
 earnest and intelligent efforts at abatement, is guilty of a barbar- 
 ism not tolerable in this age of enlightenment and progress. The 
 interest taken in the quality of public water supplies during the 
 past ten years is well shown by the work of the Royal Commission 
 on Metropolitan Water Supply, London (1893) ; by the magnifi- 
 cent and far-reaching work of the Massachusetts State Board of 
 Health, 1890, et seq., and of several important commissions upon 
 city water supply in this country and abroad ; and, finally, by the 
 independent labors of many able and patient investigators, like 
 Professor Frankland, Dr. Miquel, Dr. Prudden, and others. 
 
 The Royal Commission on Water Supply to London covered 
 more ground, and was more searching in its investigations, than 
 any similar body that has hitherto acted on behalf of a municipal 
 corporation ; and without regard to its conclusions, which may be 
 open to discussion, there can be no doubt of the great ability of 
 the commission and of the men called to give evidence before it. 
 
 The whole field of inquiry, from the available capacity of the 
 London watershed to the quality of the water which may be had 
 
INTR OD UC TION. 5 
 
 from the most perfect works for filtration, was fully covered. The 
 ablest men of England, in geology, medicine, chemistry, biology, 
 bacteriology, and sanitary and hydraulic engineering, were called 
 before the commission, and evidence was taken upon every point 
 which by any means could affect the quantity or quality of the 
 water required by the metropolis ; and minute inquiry was made 
 into the possibility of transmitting certain zymotic infectious dis- 
 eases by drinking-water. 
 
 The Hygienic Laboratory of Hamburg, so far as it relates to a 
 supervision of the quality of water supplied to the citizens, is 
 perhaps more complete than that of any other city in the world. 
 Dr. Dunbar, a former resident of St. Paul, Minn., and now a citi- 
 zen of Germany, is in charge of the laboratory ; and every facility 
 is afforded him for complete surveillance and control of the quality 
 of the city water supply. 
 
 The management of the niters, and maintenance of the quality 
 of the water supplied to the chief cities of Holland, are as care- 
 fully conducted in the interest of the public health as are the 
 boilers and pumping-engines operated in the interest of the public 
 purse. By the combined efforts of the engineers, chemists, and 
 bacteriologists connected with the water-works of Holland, the 
 water is pumped with the greatest ecomony of fuel, and is con- 
 sumed by the people with the least loss of life from water-carried 
 diseases. 
 
 The city of Manchester, Eng., realizing the value to its pros- 
 perity of an unimpeachable public water supply, has recently 
 bought a lake (Thirlmere) in County Cumberland, and much of 
 the proximate drainage ground, and conducts this water to the 
 city through conduits aggregating in length one hundred and two 
 miles. Vienna, from a city having at times typhoid fever rates 
 as high as any in Europe, by abandoning its former sources (the 
 Danube and wells), and seeking its water in the Austrian Alps, 
 has become one of the least typhoid fever infected centers in the 
 world. 
 
 The extensive labors of the Massachusetts State Board of 
 Health at its Lawrence experiment station have been guided by 
 two chief objects, one the treatment of urban sewage by practical 
 
6 THE PURIFICATION OF WATER. 
 
 methods, which will render the effluents innocuous to health, and 
 the other the development of information upon practical methods 
 of sand filtration of polluted waters. 
 
 Independent investigators have been seeking information upon 
 the exact chemical and biological character of various waters all 
 over the world. Research has been conducted along the line of 
 water transmission of disease ; and the organisms concerned in 
 the etiology of infectious disease have been patiently and carefully 
 studied. 
 
 The practical work of cities, and the scientific work of the 
 analysts, clearly point to great changes along the line of public 
 water supply. Thirty years ago the sand filters, which we now 
 find in the London water-works, were filtering water from the 
 rivers Thames and Lea, as they are now ; but no one at that time 
 suspected what these filters really were doing. The water com- 
 panies and consumers believed that the filters were making a 
 great improvement in the quality of the polluted river waters, but 
 the physics of sand filtration were at that time not written and 
 not known. 
 
 Naturally enough, processes conducted in ignorance of the 
 rationale of every step and each reaction seldom attain the high 
 efficiency which follows manipulation along lines based upon a 
 clear knowledge of all the causes operating to produce a common 
 result. And if the filters of the London water-works, as ope- 
 rated thirty years ago, failed to furnish water of a quality equal to 
 that now obtained, the fault was not in the principle of the filter, 
 but in the lack of experimental information upon the part of the 
 eminent engineers, who, like Mr. James Simpson, designed and 
 operated them. This knowledge has since been supplied by the 
 Pasteurs, the Kochs, the Franklands, and the Mills, who have each 
 in his way furnished some of the material by means of which 
 the practice of water purification has reached a firm foundation. 
 
 When Mr. James P. Kirkwood went abroad in the spring of 
 1866, to examine the works of water purification at that time in 
 use in several European cities, notably London, the subject of 
 water quality rested entirely upon the chemical tests for organic 
 matter. The filter was regarded as a fine strainer, or as Mr. Kirk- 
 
INTRODUCTION. 1 
 
 wood says,* "They (the sand filters) become indeed screens of 
 the greatest delicacy, intercepting all material impurities, not the 
 least of which are the very small fish with which all waters are 
 crowded at certain seasons." 
 
 Something smaller than fish were then held back by the 
 London filters ; this much was known, but no mention had then 
 been made of the action of bacteria in water on organic matter, 
 of the " Schmutzdecke," which in Germany is regarded at once 
 as the evidence and cause of successful sand filtration, or of the 
 action of the nitrifying organisms in converting compounds of 
 ammonia into nitrous and nitric acids. 
 
 These things were being done by the London sand filters in 
 1866, not so perfectly, perhaps, but in a measure as they are now ; 
 yet the bacteria were in the London water then as at present, but 
 no one was conducting gelatin plate cultivations, and searching in 
 drops of water for little organisms, so small in any dimension as 
 to be beneath notice, f 
 
 Organic matter in suspension in the water was being split up 
 into carbon dioxide and other gases and into nitrogenous com- 
 pounds by bacterial action ; but no one, not even Dr. Letheby, 
 mentioned it to Mr. Kirkwood upon the occasion of his visit 
 to London. The gelatinous Schmutzdecke, which Herr Piefke J 
 writes upon so ably, and argues as the very essence of successful 
 sand filtration, was being formed on the sand-beds ; but the British 
 workman, who shoveled off the upper one-half or three-quarters of 
 an inch of sand from a clogged filter bed, never noticed it. The 
 partial or complete aeration of a filter when it was temporarily 
 out of service was never suspected as a means of maintaining 
 the nitrifying organisms in the sand-bed. In short, the real ac- 
 tion of a sand filter was then unsuspected, the bed of sand being 
 considered somewhat superior to a molder's sieve for the intercep- 
 tion of suspended matter in the water. 
 
 The celebrated Dr. Letheby freely admitted " that we have 
 
 * Filtration of River Waters, by James P. Kirkwood, New York, 1869, p. 7. 
 f This term is not here used in the same sense as under observation. 
 \ Die Principien der Reinwassergewinnung vermittelst Filtration, Berlin, 1887- 
 Filtration of River Waters, by James P. Kirkwood, New York, 1869, p. 26. . 
 
8 THE PURIFICATION OF WATER. 
 
 not at the present time any absolute test for discovering organic 
 matters in water, much less the nature of these organic matters;" 
 but great as has been the progress in the chemistry, biology, 
 and bacteriology of water since Dr. Letheby penned these lines, 
 much remains to be done in applying the knowledge gained in a 
 practical way. 
 
 If it be true that by proper prophylaxis certain zymotic diseases 
 may be made to disappear, why are we so slow in adopting the 
 remedies which science and history have pointed out ? Are we 
 in doubt of the correctness of our conclusions, or are we indiffer- 
 ent to the sacrifice of human life ? 
 
 It is not the author's purpose at this time to discuss any special 
 methods of purification for polluted waters. This will be done 
 under their respective headings ; but the fact has been demon- 
 strated so often, especially abroad, that methods upon a large scale 
 can be so conducted as to command the quality of a water supply, 
 and one who opposes the purification of water supplies upon the 
 ground of impracticability must be set down as not being well 
 informed on water purification or as an enemy of the public health. 
 
 Professor Percy Frankland,* after comparing the operation of 
 the London filters for a series of years with certain deductions 
 which he had drawn from an earlier investigation of these filters, 
 stated : ' 
 
 " The importance of these results lies in their proving that in the matter 
 of sand filtration we are no longer working in the dark, but that we now know 
 the factors upon which the success of the process depends, and by attention to 
 which its efficiency may be maintained or even increased." 
 
 The tracing of disease through a sewage-polluted water may 
 be obscure to some; but if a certain source of water supply is 
 known to be polluted at some point with the organisms concerned 
 in disease, and it is further known that such organisms, or some of 
 them, can live in water for a length of time sufficient to pass from 
 the place where they enter this source of water supply to another 
 place where water is taken up for domestic uses, is it difficult to 
 conceive that some of the people who drink this water at the sec- 
 
 * Micro-organisms in Water, by Percy and Grace Frankland, London, 1894, p. 131. 
 
INTR OD UC TION. 9 
 
 ond place may take these organisms into their s'ystem and lay the 
 foundation of disease ? 
 
 The typhoid bacillus is seldom found in water, and the failure 
 to find it is too often taken as an evidence of its non-existence 
 there.* Dr. T. M. Prudden, however, in a conversation with the 
 author, very aptly disposes of this objection by stating, " If I were 
 to go down to the Battery and throw a coin into New York Har- 
 bor, do you think I could ever find it again?" The coin is there; 
 this we know because he threw it into the water, and the failure 
 to recover it cannot be taken as proof of its non-existence in the 
 bay, but as an indication of the inefficiency of our methods of 
 search. The same argument will hold good in case of failure to 
 find, among a lot of vigorous water bacteria, the typhoid or any 
 other disease germ which can sustain at best only a limited, migra- 
 tory existence in any kind of water. 
 
 It is also held by opponents of the water transmission of infec- 
 tious disease that the evidence is lacking of the actual infection 
 by this means. Of course no one sees the germ in water, and 
 therefore is never distinctly aware of taking it into the system in 
 this way ; but circumstantial evidence of the transmission of disease 
 is sometimes as potent as circumstantial evidence of crime, and 
 must be accepted accordingly. 
 
 Upon another occasion the author attempted to illustrate the 
 passage of disease germs from one point to another in water in 
 the following manner : 
 
 If we were to take an iron pipe two or three feet long, put a marble in 
 one end, tilt the pipe slightly, and make the marble appear at the other end, 
 you would say that the marble had passed through the pipe. You saw it put 
 in at one end, and in due time it appeared at the other, but you have not really 
 seen the marble passing through the pipe. The inference, however, that the 
 marble did pass through the pipe is correct ; there is no other way in which, 
 after it was put in at one end, it could reach the other. 
 
 Let the marble be replaced by the typhoid bacillus, and let the pipe be a 
 river, or a lake, or a reservoir (as at Plymouth, Penn.. in 1885). We can prove 
 by examination of the faeces of typhoid patients in the early stages of the dis- 
 ease that the Eberth germ is passing into our sewers,f and, of course, into our 
 
 * Twenty-fourth Annual Report Massachusetts State Board of Health, p. 531. 
 
 t " Report of Royal Commission on Metropolitan Water Supply," Minutes of Evidence, 
 
 UNIVERSITY 
 
10 THE PURIFICATION OF WATER. 
 
 larger sources of water supply. Eventually we will find, upon post-mortem 
 examination of persons dying in the early stages of typhoid, this same bacillus 
 in cultures made from the spleen and sometimes from the intestine. 
 
 How has it come there ? We saw it go into the sewer, and we find it in 
 the body of the typhoid victim. We know it went from the sewer to the river, 
 and we infer that the river was the carrier of the germ. We did not see it 
 passing through the water, neither did we see our marble passing through the 
 pipe. We know that the marble did go through the pipe, and I think the evi- 
 dence now before us sufficiently demonstrates that water is the carrier of the 
 typhoid bacillus from the sick to the well. 
 
 It is strange that, in spite of our exact information upon the matters which 
 convert water into sewage, we are so willing to drink this dilute mixture of filth. 
 We know sewage consists of the wastes from the household and factory, and 
 from the wash of the streets and roads, and still we drink the mixture, often 
 with no misgiving, and rarely indeed with complaint. 
 
 At the same time, if I were to take a glass of distilled water which is 
 wholly destitute of dangerous organic matter and bacteria, and in your presence 
 put into it even the slightest amount of any of the objectionable wastes which 
 constitutes sewage, there is not one person who would care to drink it. Senti- 
 ment revolts at the bare suggestion of drinking a water with which we have 
 seen filth mixed, and at the same time we swallow just such stuff when we 
 drink the water of many of our large cities. 
 
 The whole theory of water purification is based upon' the con- 
 viction that pure water cannot create a disturbance of the animal 
 system or be the cause of ill health, and that certain organic mat- 
 ter, or the products of organic matter, or organisms in water, is the 
 cause of certain disorders, or are concerned in the etiology of spe- 
 cific disease. It is not necessary for one to believe in the germ 
 theory of disease before he can become an advocate of pure water 
 supplies. Long before the ptomains and bacteria were known, 
 certain able men had pointed out that water from sources appar- 
 ently beyond the reach of pollution was more healthful to drink 
 than water which was known to be polluted. But to those who do 
 believe in the transmission of some infectious diseases by living 
 organisms, it is not difficult to perceive how sewage-polluted 
 waters may become very dangerous distributers of infection. 
 
 Pure water is held by some to be better than purified water. 
 This undoubtedly is true ; but the sources from which pure water 
 is available are so few, that it can safely be assumed if cities are to 
 have pure water, they must adopt artificial means to make it so. 
 
INTRODUCTION. 11 
 
 f ' 
 Thus filtration and sedimentation are not adopted at the present 
 
 time by any city simply to improve the appearance of water, and 
 make it more welcome to the bodily senses, but as distinct safe- 
 guards against water-borne diseases. 
 
 No one should be deceived upon the influence of sedimentation 
 or filtration of polluted waters. These means never have rendered, 
 and probably never will render, such waters pure ; but they can be 
 devised and operated in such a manner that nearly every natural 
 water can be rendered less likely to injure the human system, and 
 at a cost which will not be prohibitory to their use. A claim such 
 as is sometimes put forth, that the water from the filters of Lon- 
 don has not or cannot be the cause of typhoid fever,* cannot be 
 universally admitted. 
 
 Organisms larger than the typhoid bacillus have repeatedly 
 been detected in the filtered London water ; f and while the typhoid 
 bacillus has not been found among them, neither has it been found 
 upon careful investigation in the raw water before it has gone to 
 the filters. :f (According to Dr. G. Sims Woodhead, it has never 
 been found in any rapidly flowing river.) At the same time, with 
 greater care in operation of the filters, and with improved methods 
 of water analysis and higher standards of purity, the typhoid rates 
 of London have shown a marked decline. 
 
 Thus for the decade 1861-1870 || the annual typhoid fever 
 death rate for London was 90 per 100,000 of population. For the 
 following decade the annual typhoid fever death rate was 24 per 
 100,000 of population, and for the decade ending with 1890 the 
 annual typhoid fever death rate was 19 per 100,000 of population. 
 
 During the seven years ending Dec. 31, 1896, the average 
 annual typhoid fever death rate for London was 14.4 per 100,000 
 of population, or was then one-sixth of the rate which prevailed 
 thirty years before. This remarkable reduction in the typhoid 
 rates cannot be credited to improvements in the filters, so much as 
 
 * " Report of Royal Commission on Metropolitan Water Supply," 1893, Minutes of Evi- 
 dence, p. 404 ; also Potable Water, by Floyd Davis. New York, 1891, p. 40. 
 f Analytical Investigation of London Water Supply, 1896, p. 10. 
 | " Report Royal Commission," Mimdes of Evidence, p. 405. 
 Ibid., p. 505. 
 || Engineering Record, Oct. 27, 1894. 
 
12 THE PURIFICATION OF WATER. 
 
 to a better knowledge of how they should be operated, and to the 
 methods of water analysis developed during the past fifteen years. 
 
 Thirty years ago the London filters were operated to secure a 
 clarified water, clear water seeming at that time to mean pure 
 water, or water safe for drinking and other dietetic uses. We know 
 better now, and limpidity is no longer taken as an evidence of 
 purity in water. 
 
 When we consider that the death rate from typhoid fever has 
 been as low as three persons per 100,000 of population in Munich 
 (1892), while it has been as high as 154 persons per 100,000 of 
 population in Chicago (1891), the most obtuse must admit that 
 there is something wrong in our sanitary works or regulations 
 which will permit of a death rate from typhoid fever in any city 
 of this country fifty times as great as that of a certain city in 
 Europe. While the Munich rate is very low, still it is not excep- 
 tional, as is shown by the following rates for that city, Berlin, and 
 Vienna : 
 
 DEATHS PER 100,000 OF POPULATION FROM TYPHOID FEVER. 
 YEAR, 1890. 1891. 1892. 1893. 1894. 1895. 1896. 
 
 Munich, 8 7 3 15 2-3 3 3 
 
 Berlin, 9 10 8 9 455 
 
 Vienna, 9 6 8 7 565 
 
 Now compare these rates for the same years with those of 
 three cities of the United States. 
 
 DEATHS PER 100,000 OF POPULATION FROM TYPHOID FEVER. 
 YEAR, 1890. 1891. 1892. 1893. 1894. 1895. 1896. 
 
 Chicago, 92 154 106 45 31 32 46 
 
 Pittsburg, . . 100 100 111 56 77 61 
 
 Louisville, 88 81 72 84 72 77 45 
 
 Vienna and Munich are supplied with the purest of natural 
 waters from mountain springs, and the city of Berlin takes its 
 supply from the River Spree and Lake Tegel, the waters of both 
 being passed through artificial sand filters before they are served 
 to the consumers. 
 
 In comparing the typhoid fever rates of American and German 
 cities, perhaps some allowance should be made for the difference 
 
INTRODUCTION. 13 
 
 in habits of the populations of the respective localities. Thus 
 Munich is said to be one of the greatest beer-drinking centers of 
 the world, the consumption of this beverage having at one time 
 reached as high as one hundred and twenty-five gallons per capita 
 per annum ; and it is possible that the low typhoid fever rates 
 from the German cities may be due in part to the general absti- 
 nence of the populations from the public water for drinking pur- 
 poses. 
 
 If it be true that the low typhoid fever rates of certain cities 
 in Europe are due to the general use of beer and wine as beverages 
 instead of water, then this emphasizes the fact that a typhoid pol- 
 luted drinking-water is the principal cause of the high typhoid 
 fever rates in cities in this country, and makes it seem remarkable 
 that cities like Munich, Vienna, and The Hague, for examples, 
 where the typhoid rates are very low, and, as some people claim, 
 water is not regarded as a proper thing to drink, should pay so 
 much attention to the quality of their public water supplies. 
 
 Why should Vienna, for instance, be at such great expense to 
 bring water from the Alps, distant sixty-five miles, if it is not to 
 be used for dietetic purposes, when the water of the Danube will 
 meet every other requirement quite as well as this " Schneeberg 
 water," and can be obtained at a fraction of the cost involved in 
 the scheme of works by which that city is now supplied ? 
 
 The usual manner of introducing the typhoid germ into the 
 human system is by infected drinking-water ; therefore every city 
 should regard it as a duty to itself to see that the water distributed 
 for drinking and other dietetic purposes is not the carrier of the 
 typhoid bacillus or of the organism productive of typhoid fever. 
 
 Much has been written upon the subject of water supply and 
 the dangers of polluted waters to health, much also has been writ- 
 ten upon methods of purifying polluted waters, and doubtless much 
 remains to be written upon all these subjects ; but from the pres- 
 ent view, it can be safely stated that whenever a steady, vigorous 
 effort is made by all municipalities to supply their citizens with 
 water up to the highest standard attainable by practical means, 
 that the case and death rate from water-borne diseases will sink so 
 low as to be no longer the cause of alarm. 
 
14 THE PURIFICATION OF WATER. 
 
 CHAPTER II. 
 
 SOURCES OF PUBLIC WATER SUPPLY. 
 
 THE sources of public water supply for cities are rivers, natural 
 lakes, large impounding reservoirs, usually at elevations sufficient 
 to furnish a supply to cities by gravity ; springs, shallow dug wells, 
 often carried into the drift a depth not much in excess of the max- 
 imum suction lift of pumping machinery ; and deep driven wells. 
 Some of the latter may be artesian, and supply into large wells or 
 reservoirs from which the suction of the pumps is taken, or may 
 be connected directly with the pumps. 
 
 Of course all water supply must be derived from the rainfall, 
 whether it be taken from streams, lakes, impounding reservoirs, or 
 from springs and wells ; but the water supply of any particular lo- 
 cality may not be wholly dependent upon the local rainfall. This 
 is true where the source of supply is a river draining a large terri- 
 tory, or where it is obtained from springs or deep wells. In local- 
 ities where the outcrop or denuded rock formation is destitute of 
 soluble materials, such as lime and magnesia, the water gathered in 
 impounding reservoirs will be quite as soft as that of domestic cis- 
 tern water collected from the roof of a residence or other building. 
 
 In the limestone regions surface water, while running off to im- 
 pounding reservoirs, comes in contact with the outcrop of rock, and 
 takes up some of the lime and magnesia ; and the impounded water 
 will be harder than domestic cistern water. River water ip hardness 
 and quality will depend entirely upon the character of the water- 
 sheds from which it is derived, and the materials with which it 
 may come in contact after it has reached the channel of discharge. 
 
 Aside from the direct sewage pollution of the large rivers of 
 the world by the refuse from the civilization which is collected 
 upon their banks, there is another pollution, due to the contact of 
 the water while running off, with organic matter from various 
 
SOURCES OF PUBLIC WATER SUPPLY. 15 
 
 sources collected upon the watershed. Some of this may be car- 
 ried in solution by the runoff of rainfall into the channel, while 
 other portions may be carried along in mechanical suspension. 
 
 The objection to a polluted river or lake water is not limited 
 to the amount of sewage which it may contain. It may be posi- 
 tively objectionable from a sanitary point of view by reason of 
 organic matter, and possibly pathogenic bacteria, which may come 
 into such water from the surface drainage of the tributary water- 
 shed. If the opinion entertained by Professor E. Ray Lankester, 
 that the bacillus of typhoid fever may be an exacerbated form of 
 b. coli communis (a pathogenic germ which is known to be given 
 off in the dejecta of sheep and other domestic animals, as well as 
 of man), be confirmed by later investigation, then it is very clear 
 that the special sewage pollution of a drinking-water supply is not 
 essential for the propagation of typhoid fever, and that there will 
 be found in the organic matter now coming into streams and other 
 sources of public water supply, from the runoff of rainfall or sur- 
 face drainage, all the elements essential for the development of 
 this particular disease. 
 
 (It does not appear that there are many who share the opinion 
 with Professor Lankester that b. typhosus is an exalted form of the 
 colon bacillus, but time may demonstrate that his view is correct ; 
 and if it does, light will be shed upon some of the apparently inex- 
 plicable phenomena connected with certain epidemics of typhoid 
 fever.) 
 
 Rivers which receive the drainage of cities or towns on their 
 banks and the banks of their tributaries, even. in the absence of 
 known sewage pollution, cannot be regarded as uncontaminated 
 sources of public water supply ; and although it may be difficult to 
 show the presence of organic matter or of bacteria inimical to 
 health, still there will always be an amount of organic matter in 
 such rivers in process of decomposition, which may give rise to dis- 
 orders of the human system, even though they may not be the 
 cause of specific disease. 
 
 It is" probable that surface water can be impounded in reser- 
 voirs from watersheds at elevations so high as to avoid pollution 
 from all sources but the atmosphere ; and such water, although still 
 
16 THE PURIFICATION OF WATER. 
 
 open to the influence of decomposing organic matter found every- 
 where in nature, will always be purer than water collected in 
 rivers, lakes, and ponds on the low lands. 
 
 Sources of water supply at high elevations with a yield so large 
 as to satisfy the requirements of cities are of rare occurrence ; and 
 even in the few cases where such exist, the distance from the mu- 
 nicipality to be supplied is so great as to make the development 
 and utilization prohibitory for any but the larger cities. The city 
 of Vienna derives its supply of public water at the present time 
 altogether from large springs found in the Schneeberg, a portion of 
 the Austrian Alps, and brings this water through a conduit sixty- 
 five miles long to the city. The city of Munich obtains its water 
 supply from similiar springs in the Mangfall valley of the Bavarian 
 Alps. A few of the smaller municipalities in this country derive 
 their public water supply from springs or streams at high eleva- 
 tions in sparsely settled or wild districts. But sources of this char- 
 acter are not available by the majority of the cities of this or any 
 country ; and recourse must be had to such sources as are avail- 
 able, and these, as stated at the outset in this chapter, are rivers, 
 lakes, ponds, creeks, and dug and driven wells. 
 
 Considering rivers as constituting by far the largest source of 
 water supply for municipal corporations, it may be accepted as an 
 axiom " that no river is carrying during times of flood a water 
 which is fit for drinking purposes except it first be artificially puri- 
 fied ; " and if such rivers, in addition to the pollution which cannot 
 be avoided by the runoff of rainfall on the drainage areas, receive 
 the sewage of towns and cities, the water is undoubtedly not fit 
 for drinking and culinary uses until it has been dealt with in such 
 a- manner as to render it innocuous to health. It is frequently as- 
 serted or implied in text-books and reports on sewage and sewage 
 disposal, that the noxious properties of sewage are destroyed by 
 proper dilution ; but when sewage is the carrier of disease germs, 
 dilution cannot remove them. It will reduce the number of such 
 germs per unit of volume of the mixed sewage and water ; but the 
 germs are still there, and if taken into the system through drink- 
 ing-water may produce just as serious results to as many people 
 as if no dilution had occurred. 
 
SOURCES OF PUBLIC WATER SUPPLY. 17 
 
 Dilution of sewage undoubtedly reduces ttfe chances of any 
 single individual imbibing a fatal germ in drinking the water ; but 
 the germ itself will be just as dangerous, and in one aspect of the 
 case may be more dangerous, when imbibed. 
 
 It is well known to bacteriologists engaged in the analysis of 
 drinking-waters that the typhoid bacillus, for instance, will live 
 for the greatest length of time in a water devoid of other kinds 
 of bacteria, and will live for the least length of time in a con- 
 centrated sewage rich in the bacteria of putrefaction. The dilu- 
 tion of sewage therefore reduces the number of bacteria per unit 
 of volume of the water, and favors the vitality of the typhoid 
 bacillus. Considering that some of the bacteria can survive for 
 many days in water of single distillation containing naturally but 
 a very minute amount of organic matter, it will not be difficult to 
 apprehend the possibility of a high dilution of sewage distinctly 
 favoring the longevity of the typhoid germ ; and the theory that 
 a sewage-polluted water may be rendered safe for drinking pur- 
 poses by dilution must necessarily neglect this fact. Rivers and 
 their tributaries have for generations been the receptacles of 
 sewage, garbage, and all the wastes of civilization ; and even if the 
 water came into these channels free from any objectionable matter, 
 the practice of communities in making them the receptacles of 
 sewage would condemn the water from such sources as altogether 
 unfitted for drinking and some other uses. This, however, is well 
 recognized, not only in England and other countries of Europe, 
 but in certain portions of this country ; and steps are being taken 
 to prevent the pollution of streams by the refuse of organized 
 communities. 
 
 Considering that even in its best condition the water of a river 
 is of questionable hygienic quality, one can appreciate the impor- 
 tance of a disposal of sewage, garbage, and other wastes, in a man- 
 ner that can by no means injure sources which are drawn upon 
 for public water supply. 
 
 Dr. G. Sims Woodhead * says : 
 
 " If river water must be used, every possible precaution should be taken 
 against its being made a receptacle for unpurified sewage. It is almost im- 
 * Report Royal Commission on Metropolitan Water Supply, Appendix C, p. 491. 
 
18 THE PURIFICATION OF WATER. 
 
 possible during periods of flood to obtain it free from large quantities of sur- 
 face drainage, but it should be insisted that in ordinary weather there should 
 be no surface drainage directly into the Thames or into its tributaries." 
 
 According to Professor Baumeister,* we can safely drink a 
 sewage-polluted water when the sewage and water are mixed in 
 certain proportions, depending upon the amount of organic matter 
 in the sewage and of that previously in the water. But he depends 
 altogether on chemistry for the test of potability of the water, 
 when it is well known that chemistry is powerless to reveal infec- 
 tious properties or bacteria in the water. He states upon German 
 authority that water may carry 2^ grains of organic matter to the 
 gallon and still be potable. Looking at the question from a chem- 
 ical standpoint this may be true, but from a sanitary standpoint 
 any organic matter in drinking water known to be from a sewage 
 source will render such water unsafe for drinking. 
 
 The statement by Professor Baumeister that sewage containing 
 29.2 grains of organic matter to the gallon may be mixed with 
 river water containing 1.2 grains of organic matter to the gallon, 
 in the ratio of 23 gallons of water to 1 gallon of sewage, and the 
 mixture be safe for drinking purposes, seems to me to be very 
 dangerous, because the sewage may contain the germs of typhoid 
 fever which no amount of dilution can eliminate. Moreover, 
 advice like this, instead of promoting the purity of water supplies 
 and the public health, is calculated to injure both. 
 
 It is but fair to state that Professor Baumeister is looking at 
 the matter solely from the standpoint of sewage disposal into 
 running streams ; and it is doubtful if the selfish motive of ridding a 
 community of sewage by generally the easiest and cheapest method 
 should be allowed to prevail, when certain disaster to those who 
 may draw their drinking-water from the stream below is bound 
 to follow. 
 
 In another paragraph Professor Baumeister says : 
 
 " The objection may be raised to these computations [relating to sewage 
 dilution] that the limiting amounts of organic matter in potable water was not 
 fixed under a supposition that a part of it was human excrement." 
 
 * " The Quality of Water Supplies," by the author, Transactions American Society of Civil 
 Engineers, vol. xxxii., p. 149 et seq. 
 
SOURCES OF PUBLIC WATER SUPPLY. 19 
 
 But all sewers receive some human dejecta and this at times 
 may contain disease germs, and these germs, mixing with a so- 
 called potable water, are dangerous. Adapting to our purpose the 
 memorable words of Mr. Lincoln, any water likely to be adopted 
 for drinking purposes may be safe to all people at some times, 
 it may be safe to some people at all times, but it may not be safe 
 to all people at all times ; and the protection of those who may at 
 some time be susceptible to its deleterious influences should warn 
 us against the use of any drinking-water known to contain organic 
 matter from a sewage source. 
 
 What has been said by way of objection to rivers and their 
 tributaries as sources of water supply will apply to lakes which 
 receive the drainage of rivers and the runoff of large watersheds, 
 with the reservation, that in large bodies of water the reduction 
 of organic matter by subsidence and bacterial action will proceed 
 with a more regular rate than in rivers. In support of this prop- 
 osition, the experiments of Dr. Miquel upon water taken from the 
 River Seine at a point below the outfall of some of the larger 
 sewers of Paris indicate that polluted water at rest through a 
 long period of time is sufficient to reduce the organic matter to 
 harmless nitrates and nitrites, and remove the bacteria altogether ; 
 these (as organic matter) probably disappearing as gases, or form- 
 ing a part of the residual compounds of nitric or nitrous acid 
 with the inorganic bases. 
 
 The time required, however, according to Dr. Miquel, is very 
 great ; and excepting in cases of very large, deep bodies of water 
 no such reduction can be expected. 
 
 The experiments of Dr. Miquel have not taken into considera- 
 tion the seasonal disturbances of large bodies of water, which 
 occur in the spring and autumn, and are due to the difference of 
 temperatures of the layers of water at the top and bottom of 
 large lakes and reservoirs, which is calculated to bring organic 
 matter from the bottom of such bodies of water, and distribute 
 it with more or less uniformity throughout all the layers, from the 
 top to the bottom. Aside from these seasonal disturbances', the 
 fact is very well established, that there is a species of purification 
 going on steadily in all large bodies of water, which if uninter- 
 
20 THE PURIFICATION OF WATER. 
 
 rupted by the accession of fresh organic matter from the runoff 
 of rainfall and the discharge from sewage-polluted rivers and 
 streams, such water would eventually become absolutely pure. 
 This of course assumes a regimen for lakes and impounding reser- 
 voirs as well as the lesser quiescent bodies of water which is not 
 found in practice, and in the author's opinion, the time has ar- 
 rived to treat such sources of public water supply in the same 
 manner as we are preparing to treat the water of polluted rivers ; 
 for the same agencies which are operating to increase the natural 
 contamination of river waters are also operating to produce a 
 contamination of our lake and impounded waters, the difference 
 being more in degree than in kind of pollution. 
 
 In his testimony before the Royal Commission on Metropolitan 
 Water Supply, Dr. William Odling,* one of the official analysts 
 of the water supplied by the London companies, expressed the 
 opinion that no river or lake water was potable until after filtra- 
 tion, and laid particular stress upon the necessity of filtering lake 
 water before it was used for dietetic purposes. 
 
 This fact is well illustrated in the experience of the city of 
 Zurich, which takes its water from Lake Zurich, a large body of 
 -water at high elevation in the Swiss Alps, and supposed for many 
 years to be an ideal source of public water supply ; yet it is well 
 known that alarming typhoid fever rates have been traced to the 
 polluted water of the lake, and for a number of years no water has 
 been taken by that city until it has first been passed through a 
 system of sand filters. It is possible that no natural water supply 
 in the United States is superior to that of Lake Zurich, lying as 
 it does far above the usual sources of contamination ; but despite 
 this fact, we find that even this body of water is not located high 
 enough, or far enough away from the habitation of man, to insure 
 its purity through all time ; and if it be essential to filter the 
 water of Lake Zurich before it is delivered for drinking and other 
 domestic purposes, why should it not be so with any natural body 
 of water now used as a source of public supply in this country ? 
 
 The commonly accepted opinion that sewage-polluted streams 
 are capable of self-purification by flow through a reasonable dis- 
 
 * Minutes of Evidence, p. 388. 
 
SOURCES OF PUBLIC WATER SUPPLY. 21 
 
 . 
 
 tance can no longer be seriously entertained. The Sixth Report 
 of the Rivers Pollution Commission of England contains the con- 
 clusion " that there is no river in the United Kingdom which is 
 long enough to purify itself of sewage received at its source ; " and 
 it might have added, nor in any other country where fresh acces- 
 sions of sewage are being constantly received by rivers from cities 
 on their banks. 
 
 It is often held that aeration of polluted waters has a beneficial 
 effect on their quality. This theory, however, is successfully 
 disputed by the experiments of Dr. T. M. Drown for the Massa- 
 chusetts State Board of Health. These experiments show no 
 oxydizing effect of aeration on the suspended organic matter in 
 polluted waters, and a water must be very heavily charged with 
 sewage before the dissolved oxygen per unit of volume of the 
 water becomes so low as to have an injurious effect on the bac- 
 teria concerned in the destruction of organic matter. Aeration 
 may impart "life," as it were, to water; but it cannot be said to 
 have any marked influence on its quality. 
 
 It is probable that the self-purification of rivers occurs in the 
 same way as in lakes, by subsidence of the heavier organic matter, 
 and by the destructive action of the bacteria ; and these effects, as 
 shown by the experiments of Dr. Miquel on the water of the River 
 Seine, require considerable time, and are probably assisted by a 
 quiescent state of the water, two conditions not consistent with 
 rivers of steep or moderate slope, and exposed from point to point 
 in their course to renewals of organic matter from sewage and 
 drainage sources. 
 
 Certain eminent investigators still hold to the opinion that self- 
 purification of polluted streams really occurs, and that this, com- 
 bined with dilution of the sewage by accession of fresh water, will 
 be sufficient to purify a contaminated water until it is fit for drink- 
 ing. But no reliance can be placed on self-purification ; and if cities 
 are to have a satisfactory drinking-water from a source of known 
 pollution, it must be made satisfactory by some artificial means. 
 
 Impounding reservoirs, such as constitute the sources of sup- 
 ply for New York and Liverpool, are not open to the same degree 
 of pollution as lakes and rivers ; but excepting the drainage-ground 
 
22 THE PURIFICATION OF WATER. 
 
 of such sources is laid waste, and rendered free from all animal 
 influences, even such water cannot be regarded as complying with 
 the highest standard of hygiene. 
 
 Dug wells sunk a short distance in the drift are open to pollu- 
 tion from surface drainage ; and such should never be adopted for 
 a public supply except they be far removed from human habita- 
 tion, and then only when the materials through which they are 
 dug contains a thick stratum of impervious clay overlying the 
 water-bearing sand or gravel. Wells of this character usually are 
 limited in depth to the suction lift of pumps, and intercept water 
 only in the upper layers of the soil. 
 
 Sometimes shallow wells intercept veins of water gathered on 
 distant and higher watersheds, and the water may have been sub- 
 jected to efficient natural purification before it reaches the well. 
 In such cases, if the materials of the drift and the manner of 
 constructing the well are such as to effectually exclude all local 
 surface drainage, the water may be of high quality and altogether 
 safe. It is not an easy matter, however, to determine from what 
 source intercepted ground water has come ; and shallow well water 
 should no't be used for public supply until repeated bacterial and 
 chemical tests, through a reasonable length of time, have shown 
 no possible pollution by sewage or local surface drainage. 
 
 If it be true that the typhoid fever death rates in any large 
 community is a reliable index of the quality of the public water 
 supply, then we are bound to admit that our great lakes (sewage 
 polluted as they are, especially in the neighborhood of such cities) 
 cannot be accepted as satisfactory sources of public water supply, 
 excepting the water be subjected to careful filtration before it is 
 supplied to the consumers. 
 
 TYPHOID FEVER DEATH RATES PER 1OO,OOO OF POPULATION LIVING. 
 CITIES USING LAKE WATER. 
 
 Chicago, (average for seven years ending December 1896), 71 
 
 Milwaukee, 29 
 
 Detroit, " " " " 30 
 
 Cleveland, 46 
 
 Buffalo, " sup " 34 
 
 Average, 42 
 
SOURCES OF PUBLIC WATER SUPPLY. 23 
 
 CITIES USING RIVER WATER. 
 
 Pittsburg, (average for seven years ending December, 1896), 84 
 
 Philadelphia, " " " " " 45 
 
 Cincinnati, " " " " " " 49 
 
 Louisville, " " " " " " " 74 
 
 St. Louis, " " " " " " 39 
 
 Average, 58 
 
 CITIES USING FILTERED RIVER WATER. 
 
 London, (average for seven years ending December, 1896), 14.4 
 
 Berlin, " " " " " " " 7.1 
 
 Rotterdam, " " " 5.7 
 
 Hamburg, " four " " 9.7 * 
 
 Hamburg, " " three " " " " 7.0 
 
 Altona, " " six " " 1895 26.8 
 
 Average, 12.2 
 
 In addition to the causes of pollution of river and lake waters 
 previously mentioned, certain objectionable properties are some- 
 times imparted to the water by the subsoil drainage from irrigated 
 and fertilized land. Thus the salts in phosphates and other ferti- 
 lizers, and the ammonias from land laid with stable compost, are 
 taken up by the water percolating through the arable ground, and 
 eventually find their way by lateral movement through the soil 
 into sources of water supply. While the simple addition of organic 
 matter to water by this cause may never be very objectionable in 
 itself, there is an objection to imparting properties to water which 
 may encourage the growth and development of some of the patho- 
 genic bacteria, and the increase of the alkalinity of water has 
 already been pointed out, at least in one instance, as the cause of 
 the rapid development of the cholera bacillus. f A professor of 
 chemistry in one of our Western universities has stated to the 
 author, that from his investigations typhoid fever seems to be 
 most persistent in those districts where the water is abnormally 
 high in nitrates and nitrites, and it is altogether probable that the 
 subsoil drainage of farm lands is concerned in maintaining this 
 condition of nitrates and nitrites in certain water sources which 
 are drawn upon for domestic supply. 
 
 * Filters put in service, May, 1893. 
 
 t Micro Organisms in Water, by P. F. & G. C. Frankland, London, 1894, p. 300 (Ham- 
 burg Epidemic, 1892). 
 
24 THE PURIFICATION OF WATER. 
 
 The author does not propose at this time to discuss the influ- 
 ence of nitrates or nitrites on the vitality of the typhoid bacillus, 
 and will simply suggest that what has hitherto been regarded as a 
 matter of no consequence in connection with a public water sup- 
 ply may become, in the light of future developments on the biology 
 of this germ, a question of grave concern. Thousands of acres of 
 farm land are to-day being annually treated with natural and arti- 
 ficial fertilizers, and the subsoil water from such land is going into 
 some of our sources of water supply with possibly no advantage to 
 the water. If the opinion now held by some investigators be con- 
 firmed by later experience, that the addition to water of certain 
 salts from these fertilizers is favorable to the growth of the ty- 
 phoid bacillus, then a new and difficult problem will be pre- 
 sented in connection with the other and well-recognized sources 
 of pollution by surface drainage and urban sewage. 
 
 It is altogether feasible to provide against the direct contami- 
 nation of water supplies from sewage by requiring all communities 
 to treat this in such manner that the effluent shall conform to a 
 given standard of hygiene before it is permitted to go into our 
 water courses, lakes, and ponds ; but the objection to surface and 
 subsoil drainage cannot be so easily disposed of. In the light of 
 the present information upon the subject, we are safe in assuming 
 that any dangers to our sources of water supply from these causes 
 must be met by treatment of the water after such pollution has 
 occurred, rather than by efforts to prevent pollution ; and if the 
 theory and operation of sand filtration be accepted as established 
 conditions, and not as propositions still to be proven, we can as- 
 sume that the filtrate may be brought to any practical standard of 
 hygiene without regard to the quality of the water from which it 
 is obtained. 
 
 Filtration to be successful must be able to meet all the vary- 
 ing conditions of any water, and render a filtrate which will be 
 substantially unvarying in quality. While the quality of the water 
 applied to the filter may, and in many cases will vary between 
 wide limits, the quality of the filtered water must be practically 
 uniform. The London standard of bacterial quality of the filtered 
 water is one hundred bacteria per cubic centimeter of the filtrate ; 
 
SOURCES OF PUBLIC WATER SUPPLY. 25 
 
 and while the counts are usually much lower than this, under no 
 condition can the filtered water show more than this number with- 
 out passing the limits there assigned for potable water. 
 
 The London standard is thus not based upon the relation of 
 the numbers of bacteria in the filtrate to the numbers of bacteria 
 in the unfiltered water, but is an absolute standard, to which the 
 filtered water must conform without regard to the bacterial con- 
 dition of the water as it comes from its natural source. The 
 standard of filtered water, like all standards, is an arbitrary one, 
 and is fixed upon the judgment of men best informed upon the 
 subject ; and as standards of quality for any substance are rarely 
 placed beyond the reach of practical methods, it is reasonable to 
 infer that with increased experience and knowledge of filtration, 
 and with improved results from the application of research and 
 experiment, that the standard of water quality will be placed higher 
 and higher, until the limit of practical methods is attained. 
 
 That one hundred bacteria or colonies per cubic centimeter of 
 filtered water is not a rare or difficult achievement is well attested 
 by the operation of the Chelsea filters, which according to Dr. E. 
 Frankland, the official analyst of the water supplied by the London 
 companies, furnished a filtrate that contained for the year 1896, 
 omitting the month of June, an average of 21 colonies of bacteria 
 per cubic centimeter of water, the numbers being as high as 55 in 
 December and as low as 2 in September, while the river water at 
 Hampton Court, the point of intake for the Chelsea water company, 
 contained so few as 1,740 bacteria per cubic centimeter in August, 
 and as many as 160,000 bacteria per cubic centimeter in December 
 of that year. 
 
 No operation suffers by care in its performance ; and to the 
 caution as well as skill displayed in the operation of the London 
 filters is due the low numbers of bacteria in the filtered water, 
 and the low typhoid fever rates of that metropolis. 
 
 When failures have been recorded in the filtration of public 
 water supplies, it can be set down as being due to ignorance or 
 carelessness in proportioning the filters, or to gross mismanage- 
 ment in their operation. Many attempts have been made to pass 
 water through sand filters at rates which were not only beyond 
 
26 THE PURIFICATION OF WATER. 
 
 all precedent, but beyond reason. Thus a certain water company, 
 which is now supplying a city of over 150,000 population east of the 
 Rocky Mountains, has attempted to filter a polluted water at the 
 rate of nearly 200,000,000 gallons per acre per day ; a rate one 
 hundred times greater than that for the London filters, and has 
 assumed that this water was fit to go to its consumers, and be 
 used for drinking and other dietetic purposes. The vertical rate of 
 filtration in the London and most of the European works seldom 
 exceeds 8 to 10 feet per day of twenty-four hours, while the esti- 
 mated rate for this improved system of filtration in the Western 
 city was 600 feet per day, or 5 inches per minute.* Natural fil- 
 tration through the pervious materials of the drift is variously stated 
 to occur at rates of 7 to 40 feet per day of twenty four hours. 
 
 Under circumstances like these it is not surprising that the 
 water was really not filtered at all, and went through the mains to 
 the consumers with no actual improvement in its hygienic quality. 
 
 The typhoid fever rates for that city were abnormally high for 
 the last six months of the past year (1896), and the health officials 
 very justly charged the unusual rates to this sham filtration. 
 
 In another Western city an improved natural filter was recently 
 started to operate at rates of 22,000,000 to 44,000,000 gallons per 
 acre per day, with very satisfactory results, according to report of the 
 designer. No analysis of the water before and after it passed this 
 filter, nor records of the influence of such water filtration on the 
 health of the consumers, are available by the author ; but it cannot 
 be doubted that filtration under these conditions is really no filtra- 
 tion at all, and is calculated to hinder rather than encourage proper 
 efforts in the direction of water purification by practical methods. 
 
 If the bacterial contents of a water is a fair test of quality, then 
 driven wells sunk to moderate depths in the drift do not always 
 intercept thoroughly filtered water. Professor Sedgwick, of the 
 Massachusetts State Board of Health, has tested the water of a 
 number of driven wells in the vicinity of Boston, with bacterial 
 counts as high as 1,376 per cubic centimeter, while the water from 
 other driven wells was shown to contain so few as 30 bacteria per 
 cubic centimeter. 
 
 * Transactions American Society of Civil Engineers, vol. xxxi., p. 159. 
 
SOURCES OF PUBLIC WATER SUPPLY. 27 
 
 The author's tests have shown certain drivep wells to supply 
 water containing from 2 to 4 bacteria per cubic centimeter, while 
 other wells have shown as many as 1,060 bacteria per cubic cen- 
 timeter. When chemical analyses have been made by the author 
 contemporaneous with the bacterial tests, the higher counts of 
 bacteria in driven well waters are usually accompanied by evi- 
 dences of organic matter in the water. 
 
 High numbers of bacteria in driven well water is sometimes 
 said to be due to the condition of the casing-pipe rather than to 
 the water. But this scarcely can be correct ; the manner of driving 
 tube wells and the condition of the casing are quite alike in all sit- 
 uations, and counts of bacteria per cubic centimeter of the water 
 as widely separated as 2 to 1,400 cannot be satisfactorily ac- 
 counted for by growths on the walls of the pipe. 
 
 Within the author's practice he has seen no reason to suspect 
 any variation in the condition of the interior surfaces of the iron 
 casings, while great variations in the bacterial counts of driven well 
 water have bean recorded. It is quite probable, even with foul 
 casing-pipes, that the continuous passage of water of low bacterial 
 contents over the iron would reduce the bacteria to the kinds and 
 numbers of those naturally in the water ; * and since tests of the 
 water for bacteria are usually made after long pumping of such 
 wells, it seems unreasonable to charge high numbers of bacteria 
 in water from tube wells to the growth of species in the organic 
 matter supposed to be on the interior surface of the pipes. 
 
 Natural nitration through the materials of the drift must de- 
 pend (like artificial filtration through prepared beds of sand) upon 
 several factors, chief of which are the thickness of the layers, and 
 size of the grains of sand and gravel, through which the water passes 
 to the lower levels, where it is collected in reservoirs and pockets, 
 or intercepted by strata through which vadose currents are passing. 
 
 It cannot be assumed without analysis that natural filtration 
 always produces pure water. In some examples it doubtless does ; 
 in others it may not. If the pervious materials of the drift are 
 quite porous, allowing high rates of vertical percolation, it is possible, 
 indeed probable, that such water, if originally polluted, will still be 
 
 * Practical Bacteriology, Dr. W. Migula, London, 1893, pp. 151, 166. 
 
28 THE PURIFICATION OF WATER. 
 
 polluted at considerable depths. Beds of coarse gravel below the 
 lower levels of the ground water probably have no influence on the 
 quality of the water passing through them, no straining effect can 
 be expected, nor is the author aware of any biologic action taking 
 place in deep-seated strata of pervious materials. 
 
 Dr. Rosenau * of the U. S. M. H. S., during November of 1895, 
 made a very exhaustive examination of the water supplied to the 
 city of San Francisco, and found unmistakable evidence of the 
 presence of the colon bacillus and b. proteus vulgatis in the San 
 Andreas and Pilarcitos waters, and evidence of the proteus variety 
 in the water from the Crystal Springs Reservoir. 
 
 The Visitacion water, from a series of wells 130 to 180 feet 
 deep in the sand and gravel, contained the colon bacillus. Con- 
 cerning these bacteria Dr. Rosenau says : 
 
 " The presence of the proteus indicates fermenting processes, doubtless 
 the decomposition of organic matter in the water. This organism is one of the 
 most common and widely distributed putrefactive bacteria. 
 
 " The colon bacillus is an intestinal organism, and its presence in the 
 water means contamination with alvine discharges, either of man or the lower 
 animals." 
 
 In the light of what has been said on the inefficiency of natural 
 nitration in certain localities, the discovery of b. coli communis in 
 water from the wells of the Visitacion water-works possesses es- 
 pecial significance. The presence of this bacillus in water is an 
 index of sewage pollution, either from man or animals ; and the 
 evidence of sewage pollution at this depth (130 to 180 feet) 
 clearly demonstrates that natural filtration cannot be relied upon 
 in all localities or at all times. 
 
 The water from a well sunk in a sand-bar in the Ohio River 
 near the city of Cincinnati, at a depth of 77 feet, contained a 
 putrefactive bacterium resembling b. proteus vulgaris, which lique- 
 fied 10 per cent gelatin in a cool cupboard within two days.f The 
 water from the Ohio River had a hardness at this time of 2.18 
 to 2.40 parts per 100,000 parts of water ; while the water from 
 
 * Public Health Reports, Washington, D.C., April 10, 1896. 
 
 t Report of Engineer Commission on Extension and Betterment of Cincinnati Water 
 Works, 189(3, p. 23. 
 
SOURCES OF PUBLIC WATER SUPPLY. 29 
 
 the bottom of the sand-bar had a hardness of 12 to 13 parts per 
 100,000 parts of water, indicating ground water not well purified 
 by percolation through the pervious materials of the drift. 
 
 It is fortunate, however, that the ground water generally adopted 
 for public water supply is gathered originally on suburban or unim- 
 proved land, the runoff of which at its worst is never polluted with 
 city sewage, as are our rivers and some other bodies of water ; and 
 such water is infinitely less liable to contain the bacteria of disease 
 communicable by drinking-water. 
 
 Such waters, while of higher purity than lake and river waters, 
 are not always to be accepted as indices of the efficiency of natu- 
 ral filtration, but as waters which never were seriously polluted. 
 
 The distrust of all natural sources of water supply, excepting 
 deep wells, and springs at high elevations, by many of the European 
 authorities, has given a strong impetus to filtration -of water in 
 foreign cities. "With the exception of mountain springs * such as 
 supply Vienna and Munich, or carefully planned works for ground 
 water such as supply Dresden, or deep well water from the chalk 
 strata such as supplies the Kent district of London, the foreign 
 engineers seem to regard nearly all other of the natural sources 
 of water supply as open to suspicion." 
 
 Certain standards of quality in articles of diet are recognized 
 the world over, and even the poorer grades of food materials are 
 required by law to be of a quality that will cause no injury to 
 health. All civilized nations insist upon absolute immunity from 
 disease through articles of diet ; and why should people be less 
 concerned about the quality of their domestic water supplies than 
 they are about the quality of articles of food ? No other sub- 
 stance enters so largely into the support of the animal system ; 
 and the same care and safeguards which are applied to the ordi- 
 nary articles of food should be applied to drinking-water, and water 
 for most of the domestic uses. 
 
 The objection to polluted water is not so much to the organic 
 matter which it may contain, as it is to the possibility of the pres- 
 ence of some of the bacteria concerned in the production of disease. 
 
 * The Water Supplies of Eight Cities in Relation to Typhoid Fever Rates, by the author, 
 Chicago, 1896. 
 
30 THE PURIFICATION OF WATER. 
 
 At the present time some 23 of the pathogenic organisms have 
 been found in water or sewage, among which are the germs of 
 typhoid fever and cholera. 
 
 The latter being a disease not indigenous to this country, and 
 rarely coming even by importation, it is sufficient to consider the 
 typhoid bacillus as the special object to be avoided in selecting 
 sources of water supply, or to be restrained by methods of purifi- 
 cation of polluted waters. 
 
 While other pathogenic organisms may be imbibed through 
 drinking-water, or be taken into the system through some other 
 form of contact with water, and set up processes which lead to 
 disease, the proof of this is still lacking ; and the distinguishing 
 purpose of pure or purified water is the reduction of the typhoid 
 fever rates. Moreover, the use of a naturally pure water, and the 
 processes resorted to for the purification of polluted waters, will 
 probably have the same influence on all other water-borne patho- 
 genic organisms as on the typhoid bacillus ; and remedies which 
 will be successful in excluding this one germ from our domestic 
 waters will (so far as we now know) operate with equal force 
 against all other water-carried disease germs. 
 
 Before discussing the probability of typhoid fever infection by 
 public water supplies, it may be well to remark that other dis- 
 orders of the animal system may be traced to certain inorganic 
 matters in water. Thus waters high in lime or other bases are 
 not the best for continuous use as a drinking-water. Certain of 
 these minerals may, in very limited quantities and at times, be of 
 advantage to the animal system ; but the continuous use of a water 
 high in mineral contents is recommended by physicians only in 
 special cases, and to correct certain disorders or symptoms to 
 which such waters, or rather their mineral contents, are fitted. 
 
 In early life that part of the human system which is intended 
 to eliminate the excess of salts in water and food is very powerful, 
 and capable of a large amount of daily work without injury. As 
 we grow older, and especially in advanced life, this part of the 
 system can be easily overworked ; and when it is, the blood will 
 contain an abnormal amount of these salts or their acid products 
 which lead to very serious results. The excess of salts often is 
 
SOURCES OF PUBLIC WATER SUPPLY, 81 
 
 deposited in the capillaries and other blood vessels where the cir- 
 culation is sluggish, rendering them brittle and easy of rupture bv 
 shocks or vascular pressure. 
 
 Embolisms, apoplexy, and paralysis may be traced to this de- 
 posit of lime or some other base (in excess in the blood), which 
 impedes the movement of the fluid through the vascular system, 
 and produces stresses in some of the more delicate vessels or cap- 
 illaries, which they are unable to resist. This objection to what 
 are usually termed hard waters for drinking-purposes may be very 
 refined, and too remote for practical consideration in the light of 
 the more pronounced objection to the sewage pollution of waters ; 
 but it is worthy of thought, and continued study of this aspect 
 of drinking-waters may verify the author's opinion, that a soft, 
 pure drinking-water is better for the human system than a hard, 
 pure drinking-water. 
 
 While there is a popular sentiment against the use of a water 
 known to be sewage polluted, this is neither as strong nor as well 
 grounded as it must be to secure reforms in the water supplies of 
 many of our cities. When people come to understand that disease 
 and death lurk in sewage-polluted waters, and that to drink such 
 waters, or permit others to drink them, is an invitation to suffering 
 and -loss of life, then communities will demand remedies for the 
 evils which in many instances are now but vaguely supposed to 
 exist. 
 
 There are several well recognized tests of the quality of a water 
 supply : 
 
 First, the test by chemical methods, which measures the 
 amounts of nitrogenous organic matter in water as ammonias, the 
 chlorides of sodium and potassium as chlorine, the reduction of 
 nitrogenous matters to nitric and nitrous acids, as nitrates and 
 nitrites, and finally determines the presence of minerals, as arsenic, 
 copper, etc., which may be in quantity sufficient to make a given 
 water supply dangerous to health. Chemistry divides up the dis- 
 solved and suspended matters, and indicates the nature and amount 
 of each. 
 
 Second, the test by biological methods, which deals exclu- 
 sively with the number and kinds of organisms present, and their 
 
32 THE PURIFICATION OF WATER. 
 
 probable origin in the water, and more directly and certainly 
 than chemistry determines the fitness of water for domestic 
 uses. 
 
 Finally, the quality of a given water supply may be roughly 
 but effectually determined by allowing the people to use it, and 
 noting its influence on their health. This method prevails in 
 nearly all the cities of this country ; and a comparison of the 
 typhoid fever rates from these with the rates of other cities where 
 the quality of water supply is the subject of careful and constant 
 supervision, clearly demonstrates the importance to every com- 
 munity of the best water which skill and money can provide. 
 
 For many years the cities of Jersey City and Newark, N.J., 
 drew their water supplies from the Passaic River at Belleville. 
 Above this point, as early as 1894, the river was receiving the 
 sewage from two hundred thousand people ; and, being subject to 
 tidal influence, some of this sewage was carried up and down past 
 the two water-works intakes, until it went with ebb-tides into 
 Newark Bay, or was deposited by subsidence on the bottom of the 
 river. At the time of the author's examination of the Jersey City 
 water (August, 18&4), some destruction of the sewage by the 
 action of the bacteria, infusoria, and other living organisms, in the 
 water evidently was going on, but at a rate too slow to have any 
 marked effect on its quality. In April of 1892 Newark abandoned 
 the Passaic River, and commenced to draw its water supply from 
 impounding reservoirs in the valley of the Pequannock River, while 
 Jersey City continued to take all or part of its water from the Pas- 
 saic River until November, 1896. 
 
 These two cities are separated principally by a large meadow or swamp.* 
 They are embraced in the same "system of electric street railroads, subject to 
 the same climatic conditions, and, excepting their sources of public water sup- 
 ply, there is no known reason why any marked difference in the typhoid fever 
 rates should exist between them." 
 
 " A comparison of the typhoid rates for the past seven years from these 
 two cities, however, furnishes important evidence that water is the carrier of 
 the typhoid bacillus, and that the typhoid death rate bears a just relation to 
 the sewage pollution of our sources of public water supply." 
 
 * Engineering Record, March 23, 1895. (Including rates for 1895-1896.) 
 
SOURCES OF PUBLIC WATER SUPPLY. 
 
 33 
 
 In the following table are given the death rates from typhoid 
 fever per 100,000 of population living : 
 
 YEAR, 
 
 Death Rate, 
 
 YEAR, 
 
 Death Rate, 
 
 Until 1892, and 
 for nearly four months 
 of that year, both cities 
 drew their water sup- 
 plies from the Passaic 
 River, at Belleville. 
 During April of 1892, 
 the supply of Newark 
 was changed to the 
 Pequannock source ; 
 while Jersey City con- 
 tinued the use of Pas- 
 saic water until No- 
 vember of 1896, when 
 the whole supply of 
 that city also was 
 obtained from the 
 Pequannock River. 
 While the influence 
 of the Pequannock 
 water is not so well 
 shown in the annual 
 records of Jersey City, 
 a study of the monthly 
 typhoid mortality for 
 1896 reveals the re- 
 markable vileness of 
 the water from the 
 old source. 
 
 JERSEY CITY, N. J. 
 1890. 1891. 1892. 1893. 1894. 
 
 91 95 53 60 76 
 
 NEWARK, N. J. 
 1890. 1891. 1892. 
 
 60 81 45 
 
 1893. 
 
 28 
 
 15 
 
 1895. 1896. 
 71 61-62 
 
 1895. 1896. 
 
 17 21 
 
34 
 
 THE PURIFICATION OF WATER. 
 
 Considering that the average death rate from typhoid fever in 
 Newark since the introduction of the Fequannock water is still 
 from two and a half to five times what it would be if supplied 
 with water like that of some of the larger cities of Europe, one 
 can understand how objectionable must be the Passaic River as a 
 source of dietetic water supply. 
 
 The average rate for Newark for 1890 and 1891 was 70.5 ; 
 while the average rate for 1892 to 1896 inclusive was 25.2, or a 
 reduction of 64.3 per cent. The average rate for Jersey City for 
 1890 and 1891 was 93 ; and the average rate for 1892 to 1896 
 inclusive was 64. 3^ a reduction of 30 per cent. 
 
 This reduction in the case of Jersey City is due to other causes 
 than the use of Pequannock water ; and allowing for the same 
 general influences in the city of Newark, the net reduction in the 
 typhoid fever rates by Pequannock water was 34.3 per cent. 
 
 The distinct influence of the Pequannock water when used in 
 Jersey City in comparison with Passaic water, is shown by the 
 following table : 
 
 JERSEY CITY, N. J. (1896). POPULATION, 187,OO8. 
 
 MONTH. 
 
 PUMPED FROM 
 PASSAIC RIVER. 
 Av. DAILY GALS. 
 
 BY GRAVITY FROM 
 PEQUANNOCK RIVER. 
 Av. DAILY GALS. 
 
 PERCENTAGE 
 OF PEQUAN- 
 NOCK WATER. 
 
 DEATHS FROM 
 TYPHOID 
 FEVER. 
 
 January, 
 
 18,100,000 
 
 6,600,000 
 
 27 
 
 28 
 
 February, 
 
 15,700,000 
 
 10,400,000 
 
 40 
 
 30 
 
 March, 
 
 14,700,000 
 
 11,500,000 
 
 44 
 
 16 
 
 April, 
 
 9,800,000 
 
 13,900,000 
 
 59 
 
 9 
 
 May, 
 
 13,200,000 
 
 13,000,000 
 
 50 
 
 6 
 
 June, 
 
 6,300,000 
 
 21,000,000 
 
 76 
 
 7 
 
 July, 
 
 5,700,000 
 
 22,400,000 
 
 80 
 
 3 
 
 August, 
 
 6,900,000 
 
 22,100,000 
 
 76 
 
 3 
 
 September, 
 
 7,900,000 
 
 19,900,000 
 
 72 
 
 3 
 
 October, 
 
 3,500,000 
 
 22,500,000 
 
 86 
 
 4 
 
 November, 
 
 . . . 
 
 25,500,000 
 
 100 
 
 1 
 
 December, 
 
 . . . 
 
 28,400,000 
 
 100 
 
 5 
 
 Of the 115 deaths from typhoid for the year (1896), 74 oc- 
 curred during the first three months, for which time the average 
 proportion of Pequannock water was 39 per cent. 
 
 On comparing January and February, when 67 per cent of the 
 
SOURCES OF PUBLIC WATER SUPPLY. 
 
 35 
 
 water supply was drawn from the Passaic River, with the months 
 of November and December, when all . the water was from the 
 Pequannock River, the reduction in the typhoid rates was nearly 
 90 per cent. 
 
 From the following table it appears that the typhoid death rate 
 is generally higher in Jersey City for the months of October, 
 November, and December, than for the months of January, Febru- 
 ary, and March, as it usually is for other cities ; and assuming this 
 to be true for the year 1896, then with an increased or complete 
 substitution of Pequannock water for the sewage polluted Passaic 
 water, there was a reduction of over 86 per cent in the typhoid 
 death rates. 
 
 DEATHS FROM TYPHOID FEVER, JERSEY CITY, N. J. 
 
 YEAR, 
 
 1893. 
 
 1894. 
 
 1895. 
 
 189G. 
 
 1897. 
 
 POPULATION, 
 
 175,000. 
 
 179,939. 
 
 184,173. 
 
 187,098. 
 
 190,000? 
 
 January, 
 
 12 
 
 16 
 
 12 
 
 28 
 
 1 
 
 February, 
 
 3 
 
 5 
 
 9 
 
 30 
 
 6 
 
 March, 
 
 15 
 
 1 
 
 20 
 
 16 
 
 2 
 
 April, 
 
 7 
 
 6 
 
 19 
 
 9 
 
 2 
 
 May, 
 
 2 
 
 3 
 
 9 
 
 6 
 
 2 
 
 June, 
 
 11 
 
 4 
 
 7 
 
 7 
 
 3 
 
 July, 
 
 6 
 
 10 
 
 7 
 
 3 
 
 1 
 
 August, 
 
 13 
 
 14 
 
 11 
 
 3 
 
 2 
 
 September, 
 
 8 
 
 16 
 
 6 
 
 o 
 J 
 
 2 
 
 October, 
 
 9 
 
 23 
 
 13 
 
 4 
 
 3 
 
 November, 
 
 11 
 
 7 
 
 9 
 
 1 
 
 
 
 December, 
 
 8 
 
 14 
 
 14 
 
 5 
 
 
 The average deaths from typhoid for the months of November 
 and December for the years 1893 to 1895 inclusive, during which 
 time Passaic water alone was used (corrected for population of 
 1896), were 22, compared with which months the deaths for 1896, 
 using Pequannock water alone, were 6, or the reduction was 73 
 per cent. Comparing the deaths for January and February, 1893 
 to 1896 inclusive, with the deaths for the same months of 1897, 
 the reduction of the death rate was over 76 per cent, by reason of 
 the complete substitution of Pequannock for Passaic water. The 
 water from the Pequannock was first turned into the Jersey City 
 mains Jan. 10, 1896. 
 
36 THE PURIFICATION OF WATER. 
 
 The city of Lawrence, Mass., with a population (1896) of 
 55,000, draws its water supply from the Merrimac River, after it 
 has received the sewage from Lowell, nine miles above. The city 
 of Lowell, with a population (1896) of 85,700, draws its water 
 supply partly from the Merrimac River, and partly from a system 
 of driven wells. Lawrence, however, has filtered its water since 
 September of 1893, while Lowell uses such water as is drawn 
 from the river in its natural state. 
 
 The typhoid fever death rates per 100,000 of population living, 
 for these two cities, since 1890, are shown in the following table : 
 
 YEAR, 1890. 1891. 1892. 1893. 1894. 1895. 1896. 
 
 Lowell, 158 98 90 61 55 39 42 
 
 Lawrence, 123 115 102 93 48 31 15 
 
 The average death rates for the years 1890 to 1892 inclusive, 
 before filtered water was used in Lawrence, were for Lowell 115, 
 and for Lawrence 113, or quite the same ; while for the three 
 years, 1894 to 1896 inclusive, during which time filtered water was 
 used in Lawrence, the average rates were for Lowell 45, and for 
 Lawrence 31. The percentage of reduction in the rates for Lowell 
 was 40, and for Lawrence over 72, leaving a net reduction of 32 
 per cent to be credited to the filtered water of the latter city. 
 
 This is not all that the filtered water is entitled to, according 
 to reports from Lawrence, which show that many of the mill opera- 
 tives continue to use canal water, which is unfiltered Merrimac 
 water, in defiance of the notices posted conspicuously in the mills 
 that canal water is dangerous to health, and should not be drunk ; 
 and a fairer comparison will be of the years 1890 to 1892 inclusive, 
 before the filtered water was introduced, with 1896, when the use 
 of filtered water was doubtless more general than for 1893, 1894, 
 and 1895. Upon this comparison, Lowell shows a reduction of 
 63.5 per cent on the former rates, while Lawrence shows a reduc- 
 tion of nearly 87 per cent on the former rates, or a net reduction 
 in favor of the filtered water of Lawrence of over 23 per cent. 
 
 An examination of the table indicates that some influences 
 were at work in Lowell since the filtered water was introduced in 
 Lawrence, which very materially reduced the typhoid fever rates 
 
SOURCES OF PUBLIC WATER SUPPLY. 37 
 
 of the former city ; * but whatever these influences, they were not 
 so efficient in reducing the death rates as were those of the filtered 
 water supplied to Lawrence. 
 
 Standing alone, the nitration of the polluted Merrimac River 
 water has reduced the typhoid rates for Lawrence nearly ninety 
 per cent ; and by the correction of certain errors in the design of 
 this filter, with total abstinence of the people from unfiltered 
 water, a greater reduction than this is to be expected. 
 
 The great difficulty in the way of advancing practical works 
 of water purification is the lack of proof that water is really the 
 cause of disease. A moment's reflection will convince one that 
 apart from transmission by personal contact, as in smallpox, or by 
 food, as milk, etc., all infectious diseases must be transmitted to 
 human beings from the air, the soil, or from water. The evidence 
 now that certain diseases like tuberculosis and diphtheria are due 
 to air-borne germs is very satisfactory. Similarly from the soil 
 we obtain the germs of tetanus and anthrax, and the evidence is 
 very convincing to the majority of investigators that typhoid 
 fever and cholera are almost exclusively water-carried diseases. 
 
 Dr. Edmund Rogers, an eminent physician of Denver, Col., 
 classes mountain fever with typhoid fever. Both are continuous 
 fevers, and arise from similar causes. If the fever is light, it is 
 called mountain fever ; if it becomes intense, it is called typhoid. 
 Typhoid seems to be endemic in parts of certain States where 
 mountain water constitutes the supply for potable purposes. It is 
 a mistake to assume that mountain water must be pure water, 
 where exposed, as it is in many localities, to the sewage from 
 mining-camps or other permanent or migratory settlements upon 
 the watershed above the points at which such water is taken for 
 domestic supply. Small centers of typhoid are found upon the 
 mountain slopes at all times, and these may furnish material for 
 the infection of cities dependent upon mountain water. 
 
 * After above paragraph was written, a communication to the author from Mr. R. J. 
 Thomas, superintendent of the Lowell water-works, contained the information that since 
 February, 1896, " No water has been taken from the river, direct nor through the filter [described 
 in Chapter XI.], a sufficient supply of very good water having been obtained from a system of 
 driven wells." 
 
38 THE PURIFICATION OF WATER. 
 
 A foot-note in Mr. Preller's paper on the water-works of Zurich, 
 Switzerland, contains the following statement : 
 
 " Spring water rising in the upper Alpine reaches is, in spite of its crystal- 
 line clearness, peculiarly liable to pollution by the scattered droppings of graz- 
 ing cattle, unless the whole drainage area is inclosed. Although the water 
 purifies itself to a great extent in the course of its flow, it can produce epi- 
 demics by the droppings of diseased cattle, of which cases are recorded in 
 the upper Rhine Valley, at Neuchatel, and at Appenzell." 
 
 Here is a danger to which too little attention has been given. 
 In considering the population of a given watershed no mention 
 (within the author's knowledge) has ever been made of the num- 
 ber of domestic animals, while careful enumeration is given of the 
 people per square mile. Domestic animals are not always in a 
 state of health ; and evidently any disease germs which may be in 
 the excreta of these scattered over a given watershed will be 
 washed into the streams, lakes, or reservoirs with each succeeding 
 storm. 
 
 It is not likely that any farmer would fancy having the sewage 
 from his stock discharged into his domestic well, yet the same 
 thing really occurs when the runoff of rainfall on perhaps every 
 watershed carries this same sewage from domestic animals into 
 our sources of potable water supply. 
 
 Some of the diseases of cattle and sheep, for instance, are rec- 
 ognized as diseases of man ; and while no evidence exists that these 
 are infectious by water carriage, still it is certainly very imprudent 
 to assume that no danger can exist in this direction simply be- 
 cause it has not been proven. 
 
 It is sometimes held that all typhoid fever cannot be charged 
 to impure water supplies : this may or may not be true ; but in 
 any family the only things that are common to all its members 
 are the water, the soil, and the air surrounding the premises 
 all other possible causes of disease infection are affected by the 
 personal habits of the members. Nearly all articles of diet, ex- 
 cepting water and milk (as a beverage), are sterilized by cooking 
 and baking before they are ingested ; and such articles as are not 
 sterilized are usually washed with water before they are brought 
 to the table. 
 
SOURCES OF PUBLIC WATER SUPPLY. 39 
 
 Other causes than domestic water supplies liave been shown 
 to be responsible for typhoid epidemics, but water only has been 
 shown to be the cause of our high continuous typhoid fever rates. 
 
 The investigation of epidemics of typhoid fever in isolated 
 localities has suggested that in many of these the cause must 
 have been local ; and it has been held that when no known pol- 
 lution of the water supply by domestic sewage has occurred, the 
 water supply was blameless. 
 
 A little thought upon the subject suggests that, in settlements 
 far removed from the ordinary channels of typhoid infection, the 
 same causes may be at work that we find in more populous centers. 
 
 The colon bacillus may be found in any water open to pollu- 
 tion from the excremental refuse of domestic animals ; and may it 
 not be possible that the colon bacillus from a sheep or hog, when 
 taken into the human system, becomes the active cause of typhoid 
 fever ? and if it does, is it not easy to understand how epidemics 
 can arise, even when no apparent cause may exist ? It is not 
 known to the author that any one excepting Professor Lankester 
 believes that the colon bacillus may become the typhoid bacillus ; 
 and no one but Harvey, two hundred and fifty years ago, believed 
 in the circulation of the blood. Harvey, however, was right, while 
 the others were wrong ; and Lankester may be right to-day. Many 
 steps must be taken to prove his views ; and if proven by time, 
 the cause of these isolated typhoid cases will be made very clear, 
 and water again will be shown to be the carrier of the infection. 
 
 The fact has been repeatedly shown, that certain so-called 
 pathogenic organisms have their virulence exalted by contact with 
 certain other so-called non-pathogenic organisms ; and the com- 
 bined effect of the action of the colon bacillus normal to the 
 human intestine and the colon bacillus from domestic animals 
 may be the symptoms and lesions characteristic of typhoid fever. 
 
 Proof of this is lacking, but certain epidemics seem to be 
 accountable for in no other way. 
 
40 THE PURIFICATION OF WATER. 
 
 CHAPTER III. 
 
 BACTERIAL CONTENTS OF VARIOUS WATERS. 
 
 THE great variation in the numbers of bacteria counted from 
 the same source on different dates of the same month, or upon a 
 series of plates all inoculated in the same manner at the same 
 time, has frequently been noted, and is probably due primarily to 
 the lack of uniformity in the distribution of the bacteria through- 
 out the water sample, and somewhat to the nutrient properties of 
 the media employed, and temperature of growth. 
 
 When the nutrient media are from the same solution for a 
 series of three or more plates, and the conditions in other respects 
 the same, the author has frequently found a great difference in 
 the number of bacteria from successive drops of water from the 
 same sample, which can be reconciled only upon the theory of a 
 lack of uniform distribution of the organisms in the water sample. 
 It is well known, in the case of a water sample allowed to stand 
 for a few minutes, that the number of organisms varies consider- 
 ably with the depth at which they are taken by the dropping tube, 
 the smaller number being found near the surface of the water, 
 and the larger number at the bottom. To avoid an error due to 
 depth of water when the sample is taken up for inoculation of 
 the nutrient media, it is customary to shake the bottle thoroughly 
 before it is opened and the sample taken, to distribute as well as 
 possible the organisms throughout the whole volume of water. 
 
 The number of bacteria per cubic centimeter of a water sam- 
 ple also depends upon how the inoculation is made ; whether the 
 water is taken from the collecting bottle and quickly dropped into 
 the gelatin, or is given time to permit of the bacteria settling 
 to the point of the pipette before inoculating the tube. A test for 
 the effect of gravitation of the bacteria after the sample of water 
 has been taken up in the dropping tube is given below : 
 
BACTERIAL CONTENTS OF VARIOUS WATERS. 
 
 41 
 
 PLATE I. Water taken from the beaker into the tube, and a few minutes 
 allowed for the bacteria to settle to the point of the tube before the inocula- 
 tion was made. 
 
 PLATE II. Inoculation quickly made after the sample of water was 
 taken into the dropping tube. 
 
 PLATE III. Same as PLATE II. 
 
 WATER FROM DOMESTIC CISTERN. 
 
 PLATE I. Bacteria per cubic centimeter, 1,330 
 PLATE II. " 460 
 
 PLATE III. " " 480 
 
 Tests of Ohio River water as it comes through the taps of Cin- 
 cinnati have been made by the author with the following results : 
 
 BACTERIAL CONTENTS OF OHIO RIVER WATER AS SUPPLIED TO THE 
 CITY OF CINCINNATI. 
 
 DATE. 
 
 DAYS OF GROWTH. 
 
 COLONIES PER C. C. OF WATER. 
 
 July 18, 1894, 
 
 4 on gelatin, 
 
 7,665- 9,570 
 
 " 24, " 
 
 5 " 
 
 94,050-163,000 
 
 Aug. 1, " 
 
 7 " 
 
 1,680 
 
 Oct. 4, " 
 
 4 " agar, 
 
 9,856 
 
 " 15, " 
 
 5 " . " 
 
 1,872 
 
 " 15, " 
 
 5 " gelatin, 
 
 5,820 
 
 " 29, " 
 
 5 
 
 1,760 
 
 Nov. 7, " 
 
 5 " " 
 
 2,674 
 
 Mar. 13, 1895, 
 
 6 " " 
 
 20,724 
 
 July 23, " 
 
 4 " " 
 
 2,835- 2,910 
 
 Aug. 30, " 
 
 4 i it 
 
 561 
 
 Nov. 29, " 
 
 4 tt it 
 
 8,448- 9,120 
 
 Dec. 12, " 
 
 5 tt n 
 
 2,455- 3,295 
 
 Jan. 15, 1896, 
 
 n " u 
 
 3,146- 4,825 
 
 " 22, " 
 
 5 
 
 1,248- 1,704 
 
 " 27, " 
 
 4 " 
 
 1,498- 1,701 
 
 Feb. 1, " 
 
 4 " 
 
 5,025- 5,100 
 
 " 9, " 
 
 4| " 
 
 1,596- 1,717 
 
 " 10, " 
 
 6 " " 
 
 2,030- 2,155 
 
 " 16, " 
 
 71 ,. 
 
 1,442- 1,680 
 
 " 16, " 
 
 41 
 
 1,458- 1,593 
 
 Mar. 1, "" 
 
 3f " 
 
 842- 1,446 
 
 " 2, " 
 
 5 " 
 
 1,051- 1,821 
 
 Apr. 12, " 
 
 5* 
 
 1,657- 1,883 
 
 June 29, " 
 
 4 .< 
 
 2,304- 2,832 
 
 July 4, " 
 
 4 
 
 495- 644 
 
 Dec. 11, " 
 
 4 " 
 
 10,742- 11,300 
 
 " 13, " 
 
 4 .* 
 
 6,333- 9,637 
 
42 THE PURIFICATION OF WATER. 
 
 There is nothing unusual about the bacterial contents of the 
 Ohio River water. All rivers receiving sewage, or the wash of 
 soil, contain relatively large numbers of bacteria, most of which 
 are the common water species, and concerned in the breaking up 
 of organic matter. The water supply of Cincinnati is subjected 
 to no kind of purification before it is delivered to the consumers, 
 and any objection which may exist to it before it is pumped from 
 the river still exists when it reaches the consumer. Cincinnati is 
 one of the cities of this country which has a high typhoid fever 
 rate.- 
 
 According to Mr. M. N. Baker,* who has given very serious 
 consideration to the subject of sewage disposal and water purifica- 
 tion, " When sewage-polluted water must be used, means should 
 be adopted for its purification." 
 
 With large dilution of sewage containing pathogenic organisms, 
 the chance of taking any of these into the stomach through the 
 medium of drinking-water is diminished, but the longevity of the 
 organisms is increased. In an undiluted sewage the typhoid bacil- 
 lus would probably perish within a short time. In pure water, 
 that is, water free from the energetic putrefactive organisms, the 
 typhoid bacillus would live for weeks. If other organisms be 
 absent from the water, i.e., if the water is sterile, the typhoid 
 bacillus has been known to survive for three months. f 
 
 Dr. Abbott states that no bacteria are found in deep well 
 water,J but the author's experience has been quite to the con- 
 trary ; no well water, however deep the well, has failed to contain 
 some bacteria, and some moderately deep driven wells have shown 
 considerable numbers upon bacterial test. 
 
 The examinations by Professor Sedgwick, and the table of 
 results by the author which are given on pp. 44 and 45, throw 
 some light on the bacterial contents of well water. 
 
 Certain experiments have been made to determine the effect of 
 domestic filters on Ohio River water. These filters are all sold as 
 germ-proof apparatus, and the purchaser in most instances really 
 
 * New Jersey Sanitary Association, Proceedings, 1895, p. 75. 
 
 t Water Supply for Cities, by author, University of Illinois, 1896. p. 12. 
 
 \ Principles of Bacteriology, by A. C. Abbott, M.D., Philadelphia, 1894, pp. 419-436. 
 
BACTERIAL CONTENTS OF VARIOUS WATERS. 43 
 
 believes in their efficiency in the prevention of th'e passage of bac- 
 teria. The best domestic filter is the Pasteur, with which at this 
 time nearly every person is familiar ; but even this will not restrain 
 the passage of bacteria for any length of time. Variations in the 
 porosity of the porcelain tubes will increase or diminish the rate 
 of delivery of water through the unglazed material, and correspond- 
 ingly affect the rapidity with which certain of the bacteria will 
 grow through the pores of the tubes. 
 
 BACTERIAL CONTENTS OF WATER. 
 
 PASTEUR-CHAMBERLAND FILTERS, 
 
 DAYS OF GROWTH COLONIES PER C. C. 
 
 DATE - ON GELATIN. OF WATER. 
 
 1. June 16, 1894, 7 62 
 
 Tube sterilized just before use. 
 
 2. Oct. 10, 1894, 5 580-974-1,536 
 
 This filter in restaurant, and probably not well attended to. 
 
 3. Oct. 10, 1894, 5 2 
 
 This is a new filter with freshly sterilized tubes. 
 
 4. Oct. 15, 1894, 10 4 
 
 Sample from new filter. 
 
 5. Oct. 24, 1894, 5 180-209-436 
 
 This filter is in a popular hotel, and carefully attended to. 
 
 6. Nov. 25, 1894, 19 4-8 
 
 This filter in drug-store, water used for prescription purposes. 
 
 7. May 23, 1895, 7 201-236-287 
 
 Same as No. 6, water still used for prescription purposes. 
 
 8. May 23, 1895, 7 167-182-293 
 
 Same as No. 2, tubes renewed. 
 
 Freudenrich * has made some experiments with the Pasteur 
 filter to determine the sterility of the filtrate at different dates 
 after sterilization of the tubes, and for different temperatures of 
 the room in which the filters were kept, and finds that at a tem- 
 perature of 35 C. the filter delivered sterile water at the end of 
 six days, while at a temperature of 22 C. the filtrate in some cases 
 was, and in others was not, sterile at the end of ten days. The 
 cause of one filter furnishing sterile water, and another operating 
 under the same conditions giving a filtrate containing bacteria, is 
 
 * Centralblatt fur Barter iologie, vol. xii., 1892, p. 240. 
 
44 
 
 THE PURIFICATION OF WATER. 
 
 explained by the investigator as being due to the difference m the 
 density or porosity of the tubes, and to differences in the micro- 
 organisms in the water at different times. 
 
 It has been the author's experience with water from niters of 
 this type that they never furnish absolutely sterile water ; for upon 
 a series of plates inoculated with such water, while some may re- 
 
 BACTERIAL CONTENTS OF WELL WATER, EASTERN MASSACHUSETTS. 
 (From Examinations by PROFESSOR W. T. SEDGWICK. *) 
 
 LOCATION OF WELL. 
 
 DEPTH IN 
 FEET. 
 
 COLONIES PER C. C. OF 
 WATER. 
 
 Cambridge, 
 
 103 
 
 254- 269 
 
 
 
 100 
 
 30 
 
 
 
 454 
 
 206- 214 
 
 
 
 254 
 
 135- 150 
 
 Lowell, 
 
 . . 
 
 228 
 
 <( 
 
 . . 
 
 178(Gly.Agar.) 
 
 Cambridgeport , 
 
 198 
 
 116 
 
 < 
 
 198 
 
 192- 193 
 
 (C 
 
 198 
 
 258- 262 
 
 Boston, 
 
 213 
 
 138- 139 
 
 
 
 213 
 
 130- 140 
 
 <( 
 
 213 
 
 101- 106 
 
 
 
 377 
 
 48- 54 
 
 ii 
 
 377 
 
 149- 158 
 
 Cambridgeport, 
 
 277 
 
 1,240-1,376 
 
 
 
 277 
 
 486 
 
 Boston, 
 
 130 
 
 440- 480 
 
 
 
 200 
 
 525 
 
 Roxbury, 
 
 180 
 
 57- 60 
 
 Somerville, 
 
 67 
 
 165 
 
 Roxbury, 
 
 750 
 
 38 
 
 main sterile for several days, in due time they will develop one or 
 more colonies. An entirely sterile plate he has never met with. 
 Professor Percy Frankland, in discussing this filter, says : 
 
 " It must be regarded, therefore, still as an open question whether patho- 
 genic organisms, such as typhoid bacilli, can or cannot grow through the 
 pores of the Chamberland (Pasteur) filter ; and until this question has been 
 answered in the negative, it is obvious that in using these cylinders they should 
 be frequently cleaned and sterilized." 
 
 * Twenty-sixth Annual Report Massachusetts State Board of Health, p. 435. 
 
BACTERIAL CONTENTS OF VARIOUS WATERS, 
 
 45 
 
 Not having the details of the Freudenrich tests, it is impos- 
 sible to compare the results from abroad with those obtained here. 
 Tests of Pasteur filters, in such condition as they are found in 
 hotels and restaurants, have given from 180 to 1,500 bacteria per 
 
 BACTERIAL CONTENTS OF WATER FROM DRIVEN WELLS. 
 
 (From Examinations by Author.) 
 
 DATE. 
 
 LOCATION. 
 
 GROWTH ON 
 GELATIN, 
 DAYS. 
 
 DEPTH OF 
 WELL it: 
 FEET. 
 
 COLONIES 
 
 PER C. C. 
 
 OF WATER. 
 
 REMARKS. 
 
 1895 
 
 Lebanon, O. 
 
 5 
 
 62 
 
 260-1060 
 
 Boring No. 7. 
 
 1895 
 
 Wyoming, O. 
 
 34 
 
 146 
 
 109 
 
 . . . 
 
 1895 
 
 Lebanon, O. 
 
 5 
 
 115 
 
 7 
 
 Boring No. 12. 
 
 " 
 
 
 
 8 
 
 115 
 
 4 
 
 <( 
 
 (( 
 
 
 
 11 
 
 115 
 
 8 
 
 (4 
 
 1895 
 
 St. Maryls, O. 
 
 5 
 
 265 
 
 3- 14 
 
 In lime rock. 
 
 it 
 
 " 
 
 74 
 
 265 
 
 3- 14 
 
 " 
 
 n 
 
 
 
 9| 
 
 265 
 
 6- 14 
 
 
 
 M 
 
 14 
 
 5 
 
 280 
 
 53- 67 
 
 14 
 
 1896 
 
 Dayton, Ky. 
 
 54 
 
 82 
 
 34 
 
 Dayton sandbar. 
 
 14 
 
 
 
 54 
 
 82 
 
 38 
 
 <( 
 
 
 
 it 
 
 54 
 
 82 
 
 39 
 
 " 
 
 1896 
 
 Dayton, O. 
 
 44 
 
 70 
 
 1 
 
 Pumping. 
 
 
 
 M 
 
 44 
 
 70 
 
 52 
 
 . . . 
 
 (( 
 
 II 
 
 44 
 
 70 
 
 60 
 
 Natural flow. 
 
 " 
 
 14 
 
 44 
 
 70 
 
 66 
 
 n 
 
 <( 
 
 (1 
 
 4 
 
 70 
 
 146- 149 
 
 
 
 M 
 
 (( 
 
 5 
 
 70 
 
 31- 39 
 
 After 72 hours of 
 
 
 
 
 
 
 pumping. 
 
 1896 
 
 Wyoming, O. 
 
 4 
 
 146 
 
 7 
 
 From discharge pipe 
 
 
 
 
 
 
 of pump while 
 
 
 
 
 
 
 pumping. 
 
 
 
 (i 
 
 4 
 
 146 
 
 73- 75 
 
 From tap, residence. 
 
 1897 
 
 Wyoming, O. 
 
 4 
 
 146 
 
 285- 305 
 
 ( <t 
 
 (( 
 
 " 
 
 9 
 
 146 
 
 2.6- 7.8 
 
 From discharge pipe 
 
 
 
 
 
 
 of pump. 
 
 cubic centimeter ; and the low counts of 2 to 4 bacteria per cubic 
 centimeter have been obtained from tubes not previously used. 
 
 In regard to Freudenrich's * experiments with water from the 
 Pasteur filter, unfortunately we have the results of his investiga- 
 tions without the details as to nutrient media employed, or tem- 
 
 * Fire and Water, Nov. 3, 1894. 
 
46 THE PURIFICATION OF WATER. 
 
 perature and time allowed for the colonies to develop in the 
 inoculated gelatin (or other media). If Freudenrich's inocula- 
 tions were made in the standard gelatin-peptone solution, and the 
 plates or dishes were examined, as is customary, at the end of 
 three or four days, one can understand how such plates might 
 show no growth at all ; but if the examinations be delayed for a 
 week or ten days, we should expect to find colonies of water bac- 
 teria appearing on such plates. 
 
 From his own experiments with water from new Pasteur filter- 
 tubes, the author has never failed to find colonies of water bacteria, 
 if sufficient time is given for these to develop ; and this may be 
 partly accounted for by the nutrient properties of the gelatin, and 
 the great difference in time of growth between different bacteria 
 found in water, some appearing in twenty-four hours, while others 
 require from one to three weeks to grow. If a plate is planted 
 with slow-growing bacteria, an examination of such plate at the 
 end of four days may reveal no colonies at all, and such water 
 would incorrectly be declared sterile. For, evidently, if any water 
 bacteria are found growing on the plate after being kept for two 
 or three weeks in the moist chamber at room temperature, such 
 bacteria must have been in the water when the plate was inocu- 
 lated. Of course it is assumed in such cases that the plate has 
 been carefully guarded against the introduction of adventitious 
 germs. 
 
 Stone disk and tube filters, of which quite a number of forms 
 are now being made and sold, are not to be regarded as germ 
 proof, although so labeled ; and from the treatment which they 
 receive after being introduced into a residence, they usually be- 
 come a positive menace to the health of the family dependent 
 upon them for their drinking-water. It is not denied that these 
 filters are successful in clarifying turbid waters ; but a collection 
 of sponges will do the same thing, and no one supposes that a 
 sponge filter will restrain the passage of bacteria. Water may be 
 rendered perfectly limpid by filtration ; but colorless or clear water, 
 and pure or purified water, are not the same. 
 
 Clarified water may be the carrier of pathogenic bacteria quite 
 as well as turbid water ; while turbid water, apart from the inor- 
 
BACTERIAL CONTENTS OF VARIOUS WATERS. 
 
 47 
 
 ganic substances (in solution or suspension) wliich gives it color, 
 may be very pure. Clarification and filtration are not one and the 
 same thing ; although, as a rule, properly filtered water usually is 
 colorless, unless the color is derived from peaty substances, in 
 which event the hygienic quality may be greatly improved by fil- 
 tration without removing the color. 
 
 BACTERIA IN WATER FROM DOMESTIC FILTERS. 
 
 STONE DISK FILTERS, 
 
 DATE. 
 
 DAYS OF GROWTH COLONIES PER 
 ON GELATIN. C. C. OF WATER. 
 
 1. Sept. 2, 1894, 5| 19,035-48,600 
 
 This filter in a very popular restaurant, with the legend on the menu : " The water of this 
 establishment is filtered through stone filter, and is absolutely pure." 
 
 2. Dec. 18, 1894, 5 3^859-5,733 
 
 This filter in private residence filtering cistern water. 
 
 3. July 12, 1895, 4 11,704-14,605 
 
 Same as No. 2. 
 
 4. July 20, 1895, 4 29,765 
 
 Same as Nos. 2 and 3. 
 
 5. April 12, 1896, 5 2,873-3,628 
 
 New filter in service few hours before sample was taken 
 which offers its patrons pure filtered water. 
 
 6. Dec. 9, 1896, 5 1,299-1,308 
 
 Filter in residence. 
 
 this also is in a restaurant 
 
 STONE TUBE FILTER, 
 
 7. July 17, 1895, 4 
 
 8. July 19, 1895, 6 
 
 9. July 21, 1895, 8 
 
 SAND FILTERS, 
 
 10. May 21, 1895, 4 
 
 In restaurant. 
 
 11. Dec. 17, 1895, 7 
 
 Experimental filter. 
 
 12. Jan. 1, 1896, 7 
 
 Experimental filter. 
 
 13. Sept. 20, 1896, 2 
 
 125-175 Showing effect of 
 335-440 time on development 
 410-535 of colonies. 
 
 153-176-180 Without coagulant. 
 
 30-34-36 With coagulant. 
 
 20-85 
 
 77-105 " 
 
 This sample of water reported by chemist as being free from alb. ammonia. 
 
 SAND AND CHARCOAL FILTER, 
 14. March 30, 1896, 3 
 
 Filter in residence. 
 
 2,445-2,512 
 
48 THE PURIFICATION OF WATER. 
 
 It will be noted that the number of colonies from a given 
 water sample depends upon the days which the plate is permitted 
 to grow before the count is made. Referring to the stone tube fil- 
 ter, the water from which was tested July, 1895, by prolonging the 
 growth from four to eight days, the count was increased more than 
 three times ; and all. the colonies finally found on these plates were 
 due to bacteria in the water at the time of inoculation. The same 
 increased growth due to time of inoculation is strikingly shown in 
 the following table of bacteria from spring waters, where in one 
 instance the count rose from 424 at the end of two days to 1,440 
 at the end of six days. Other illustrations of the influence of time 
 on the counts will be noted in the several tables of bacterial con- 
 tents vyf various waters. 
 
 Spring waters sold in Cincinnati for table use have given the 
 following results when tested for bacterial contents : 
 
 SPRING WATER. 
 
 (Samples collected at the source, and planted within two hours.) 
 
 DATE. SPRINGS. . DAYS OF GROWTH BACTERIA PER 
 
 ON GELATIN. C. C. OF WATER. 
 
 Sept. 2, 1895, Tallewanda, Ohio, 4 128-148 
 
 Nov. 3, " " " 4 85-402 
 
 (From Bottled Spring Water.) 
 
 Geneva bottled spring water, planted Aug. 1, 1895, gave the 
 following results : 
 
 Counted at end of 2 days, 424 colonies per c. c. of water. 
 
 u 4 \ 024 " " " " 
 
 5 1,160 " " " " 
 
 6 " 1,440 " " " " 
 
 Dr. T. M. Drown, from analyses of forty-one spring waters in 
 Massachusetts,* found : 
 
 2 springs contained 1 bacteria per c. c. of water each. 
 
 5 a u 2 " " " 
 
 7 " an average of 7 u " " 
 
 Q u u u u 16 " " " 
 
 7 u u 29 " " " 
 
 Q u u u u 70 <t << 
 
 * Twenty-third Annual Report Massachusetts State Board of Health, p. 356. 
 
BACTERIAL CONTENTS Of VARIOUS WATERS. 
 
 49 
 
 3 springs contained an average of 159 bacteria per C. C. of water each. 
 
 1 " " 259 " " " 
 
 1 446 
 
 1 u u 973 H a n 
 
 2 " " " 1,844 " " " 
 
 Professor W. T. Sedgwick * has examined the water in a number 
 of springs in the country district of southern New Hampshire with 
 the following results : 
 
 No. 
 
 DATE OF 
 ANALYSIS. 
 
 BACTERIA PER C. C. OF WATER 
 
 On Gelatin. 
 
 On Agar. 
 
 1 
 
 Nov. 29, 1894, 
 
 252-258 
 
 145-167 
 
 2 
 
 
 
 134-163 
 
 133 
 
 3 
 
 
 
 92-105 
 
 72- 79 
 
 4 
 
 
 
 95-106 
 
 89- 96 
 
 5 
 
 
 
 193-218 
 
 203-217 
 
 6 
 
 tt 
 
 43-100 
 
 36- 72 
 
 The author, in searching for b. typhosus and allied organisms 
 in the Ohio River water, has made several tests with the solution 
 proposed by Parietti for destruction of the non-pathogenic bacteria 
 in water, with the results as given in the table on page 50. The 
 influence of varying quantities of the acid solution is distinctly 
 shown by the test of Jan. 5, 1896. 
 
 The test of Aug. 7, 1895, indicates the influence of time on 
 plates inoculated with Ohio River water, after treatment of the 
 gelatin with a strong dose (Jfa c.c.) of the Parietti solution. At 
 the end of four days (the usual time of incubation of water bac- 
 teria at room temperature) no growth at all had occurred, while 
 three days later twelve colonies had appeared, and three days later 
 than this the growth had increased six times. In endeavoring to 
 reduce the number of bacteria in a water sample by dosing the 
 gelatin with the Parietti solution, considerable care must be exer- 
 cised to avoid the destruction of the typhoid bacillus if it be pres- 
 ent. It is well known that several other organisms will resist 
 larger doses of the acid solution ; and upon any plate, when search 
 is being made for b. typhosus, one must expect to find several other 
 
 * Twenty-sixth Annual Report Massachusetts State Board of Health, p. 43(5 et seq. 
 
50 
 
 THE PURIFICATION OF WATER. 
 
 bacteria, if the latter happen, as is usual, to be present in the 
 water. 
 
 It is commonly supposed that freshly fallen rain-water contains 
 few bacteria. According to Miquel (Paris), rain-water at Mont- 
 souris, in the suburbs of the city, contained 4.3 bacteria per cubic 
 centimeter, while rain-water caught in the middle of the city con- 
 tained 19 bacteria per cubic centimeter. 
 
 INFLUENCE OF PARIETTI SOLUTION ON GROWTH OF BACTERIA. 
 CINCINNATI, OHIO, TAP WATER. 
 
 DATE. 
 
 DAYS OF 
 GROWTH ON 
 GELATIN. 
 
 KIND OF WATER. 
 
 COLONIES PER 
 C. C. OF WATER. 
 
 Aug. 7, 1895, 
 
 4 
 
 Plain water, 
 
 804 
 
 Aug. 7, 
 
 4 
 
 City water in 10 c. c. of 15% gelatin 
 
 
 
 
 treated with y\ c - c - of Parietti 
 
 
 
 
 solution, 
 
 No growth. 
 
 Aug. 10, 1895, 
 
 7 
 
 * 
 
 12 
 
 Aug. 13, 
 
 10 
 
 " 
 
 72 
 
 Dec. 12, 
 
 5 
 
 Plain water, 
 
 2,455-3,295 
 
 Dec. 12, " 
 
 5 
 
 City water treated with 6 drops = 
 
 
 
 
 i c. c. of Parietti solution in 10 c. c. 
 
 
 
 
 of 10% gelatin, 
 
 25- 29 
 
 Dec. 15, 1895, 
 
 n 
 
 Plain water, 
 
 3,146-4,825 
 
 Dec. 15, " 
 
 7J 
 
 City water treated with 6 drops = 
 
 
 
 
 i c. c. of Parietti solution in 10 c. c. 
 
 
 
 
 of 10% gelatin, 
 
 59- 119 
 
 Jan. 5, 1896, 
 
 18 
 
 City water treated with 4 drops 
 
 
 
 
 ^ c. c. of Parietti solution in 10 c. c. 
 
 
 
 
 of 10% gelatin, 
 
 853 
 
 Jan. 5, 1896, 
 
 18 
 
 City water treated with 5 drops = 
 
 
 
 
 I c. c. of Parietti solution in 10 c. c. 
 
 
 
 
 of 10% gelatin, 
 
 658 
 
 Jan. 5, 1896, 
 
 18 
 
 City water treated with 6 drops = 
 
 
 
 
 i c. c. of Parietti solution in 10 c. c. 
 
 
 
 
 of 10% gelatin, 
 
 227 
 
 The author, however, finds that the numbers of bacteria in 
 rain-water depend upon the time at which the collection of the 
 sample is made. If at the beginning of a shower, the numbers 
 may be high, while after a few hours of rainfall the atmosphere 
 appears to have been washed of its bacterial contents, when the. 
 
BACTERIAL CONTENTS OF VARIOUS WATERS. 
 
 51 
 
 numbers become very low. The notes below illustrate the influ- 
 ence of time of rainfall upon the bacterial contents of the water. 
 The author's samples were collected in the suburbs of Cincinnati, 
 where the conditions were quite like those of the country. 
 
 Fresh rain-water from short shower caught in sterilized bottle placed 
 on the ground clear of trees and houses. 
 
 DATE OF TEST. DAYS OF GROWTH. COLONIES PER C. C. 
 
 July 22, 1895. 4 5,495-5,759. 
 
 A sample of rain-water collected July 28, 1895, on 4 days' 
 growth, gave 414 molds and 624 colonies per cubic centimeter. 
 
 Rain-water collected at end of twelve hours of rainfall. 
 
 DATE. DAYS OF GROWTH. COLONIES PER C, C. 
 
 July 12, 1896, 3i 15-18 
 
 July 13, 1896, 4i 54-57 
 
 As further interesting information upon the chemical quality of 
 freshly fallen rain-water, the following analyses by Dr. Thomas M. 
 Drown are quoted : 
 
 CHEMICAL ANALYSIS OF RAIN-WATER* 
 
 (Parts per 100,000.) 
 
 DATE. 
 
 STATION. 
 
 AMMONIA. 
 
 NITROGEN AS 
 
 CHLO- 
 RINE. 
 
 Free. 
 
 Albuminoid. 
 
 Nitrates. 
 
 Nitrites. 
 
 1888. 
 
 
 
 
 
 
 
 July 7, 
 
 North Andover, 
 
 .0047 
 
 .0038 
 
 .0070 
 
 .0000 
 
 . . . 
 
 Sept. 18, 
 
 <( 
 
 .0016 
 
 .0026 
 
 .0040 
 
 .0000 
 
 . . . 
 
 Sept. 12, 
 
 Lawrence, 
 
 .0298 
 
 .0024 
 
 .0000 
 
 .0000 
 
 .007 
 
 Oct. 2, 
 
 
 
 .0414 
 
 .0030 
 
 .0100 
 
 .0000 
 
 . . . 
 
 Nov. 27, 
 
 
 
 .0164 
 
 .0014 
 
 .0050 
 
 .0002 
 
 .360 
 
 1889. 
 
 
 
 
 
 
 
 May 21, 
 
 Lawrence, 
 
 .0086 
 
 .0026 
 
 .0030 
 
 .0001 
 
 .070 
 
 June 17, 
 
 Jamaica Plain, Boston, 
 
 .0564 
 
 .0152 
 
 .0180 
 
 .0004 
 
 .130 
 
 1890. 
 
 
 
 
 
 
 
 Mar. 28, 
 
 Newton Highlands, 
 
 .0154 
 
 .0034 
 
 .0050 
 
 .0001 
 
 .060 
 
 1887. 
 
 
 
 
 
 
 
 Dec. 26, 
 
 Boston (snow), 
 
 .0258, 
 
 .0038 
 
 .0030 
 
 
 
 
 
 * Massachusetts State Board of Health, 1890, Part I., " Report on Water Supply and 
 Sewerage," p. 562. 
 
52 THE PURIFICATION OF WATER. 
 
 Quite a profitable business has been established in several of 
 the larger cities in the manufacture and sale for table use of dis- 
 tilled water. A test of such water as supplied in Cincinnati gives 
 a very favorable result : 
 
 DISTILLED WATER. (Single Distillation.) 
 
 DAYS OF GROWTH BACTERIA PER C. C. 
 DATE - ON GELATIN. OF WATER. 
 
 Oct. 29, 1895, 4 30-52 
 
 Oct. 31, 1895, 6 38-80 
 
 The following water was suspected of having caused typhoid 
 fever in one of the State institutions of Ohio : 
 
 (Sample taken from tap in Superintendent's Office.) 
 
 DAYS OF GROWTH BACTERIA PER C. C. 
 DATE. 
 
 ON GELATIN. OF WATER. 
 
 Sept. 11, 1896, 3 845-862-897 
 
 Many of the above colonies were rapid liquefiers, compelling 
 the count of the dishes at end of three days to avoid complete loss 
 of test. 
 
 Same water tested Sept. 29, four days growth on gelatin. 
 
 One dish gave 239 bacteria per cubic centimeter of water. 
 
 Second dish, gelatin wholly liquefied and count impossible. 
 
 Some tests of the influence of a small Anderson Revolving Iron 
 Purifier on water from the Ohio River gave results as follows : 
 
 ANDERSON REVOLVING IRON PURIFIER. 
 (Laboratory Test of Ohio River Water.) 
 
 DAYS OF GROWTH COLONIES PER C. C. 
 
 DATE. KIND OF WATER. QN GELATIN . OF WATER . 
 
 June 29, 1896, Plain city water, 4 2,304-2,832 
 
 " " Anderson Purifier, 4 50- 90 
 
 July 4, 1896, Plain city water, 4 495- 644 
 
 " " Anderson Purifier, 4 39- 119 
 
 Reference will be made to this performance again in discussion 
 of this mode of water purification ; but at this point it should be 
 remarked, that the conditions under which the experiments were 
 conducted did not favor the best performance of the device, but 
 
BACTERIAL CONTENTS OF VARIOUS WATERS. 53 
 
 were deemed sufficient to indicate, in a rough way, how the con- 
 tact of the fragments of iron in the purifier aided in the removal, 
 by subsequent sand filtration, of the larger percentage of the 
 bacteria contained in the raw water. 
 
 Artificial ice made from distilled water is now largely sold in 
 many cities, particularly where the climate is prohibitory of the 
 collection of ice from lakes and ponds. A test of such ice, as 
 supplied in the city of Cincinnati and vicinity, is given below : 
 
 TESTS OF ARTIFICIAL ICE SUPPLIED IN CINCINNATI. 
 
 DAYS OF GROWTH COLONIES PER C. C. 
 
 ON 15,% GELATIN. OF WATER. 
 
 July 12, 1896, 3 17- 37 
 
 July 13, 1896, 4 99-105 
 
 In testing water samples for bacterial contents, in addition to 
 the usual precautions to avoid contamination from the atmosphere, 
 it is advisable to make an occasional test of the air of the work- 
 room, as a guide to the probable amount and kinds of organisms 
 which might accidentally come into a sample under observation. 
 Such tests by the author indicate a considerable variation in the 
 numbers of air germs which will fall on an open sterile gelatin 
 plate in the same room at different times. 
 
 BACTERIAL CONDITION OF AIR OF WORKROOM. 
 
 DATE. TIME OF EXPOSURE. GROWTH. MOLDS. COLONIES. 
 
 July 14, 1894, 10 minutes, 4 days, 26 69 
 
 Nov. 15, 1895, 15 9 2 43 
 
 Mar. 29, 1896, 15 7 " 2 4 
 
 Certain experiments have shown sunlight to be a powerful 
 agent in the destruction of nearly all forms of bacterial life. 
 With respect to the influence of sunlight on the bacterial life in 
 river water, Dr. E. Frankland, in his " Report on the Quality of 
 London Waters for 1895," says:* 
 
 " With regard to the effect of sunshine upon bacterial life, the interesting 
 observations of Dr. Marshall Ward leave no doubt that sunlight is a powerful 
 germicide ; but it is obvious that its potency in this respect must be greatly 
 
 * Annual Summary of Vital Statistics, London. 1895, p. 67. 
 
54 THE PURIFICATION OF WATER. 
 
 diminished, if not entirely annulled, when the solar rays have passed through 
 a stratum of water even of comparatively small thickness before they reach 
 the living organisms. By a series of ingeniously contrived experiments Mr. 
 Burgess has demonstrated the correctness of this view. A sterile bottle was 
 half filled with Thames water, and violently agitated for five minutes to insure 
 equal distribution of the organisms. Immediately afterwards a number of 
 sterile glass tubes were partially filled with this water, and sealed hermetically. 
 Three of these tubes were immediately packed in ice, and the remainder were 
 attached in duplicate, at different distances apart, to a light wire frame, which 
 was then suspended vertically in the river. The experiments were made near 
 the Grand Junction Water Company's intake, at a place favorable for the sun's 
 rays to fall on the river without any obstruction. The river was at that time 
 in a very clear condition, and contained but little suspended matter, while the 
 day was fine, although clouds obscured the sun occasionally. The tubes were 
 exposed to light in the river for four and a half hours (from 10.30 A.M. to 3 P.M., 
 on May 15, 1895). At the end of this time the tubes were packed in ice for 
 transport to the laboratory, where the cultivation was started immediately. The 
 colonies were counted on the fourth day, and yielded the following numbers : 
 
 COLONIES PER C. C. 
 
 OF WATER. 
 
 Thames water packed in ice immediately after collection, 2,127 
 
 " after exposure to sunlight for 4| hours at 
 
 surface of river, 1,140 
 
 " " after exposure to sunlight for 4| hours at 
 
 6 inches below surface of river, 1,940 
 
 " after exposure to sunlight for 4| hours at 
 
 1 foot below surface of river, 2,150 
 " " after exposure to sunlight for 4 hours at 
 
 2 feet below surface of river, 2,430 
 " after exposure to sunlight for 4^ hours at 
 
 3 feet below surface of river, 2,440 
 
 These experiments show that on the 15th of May the germici- 
 dal effect of sunlight on Thames microbes was nil at depths of one 
 foot and upwards from the surface of the water. It cannot there- 
 fore excite surprise that the effect of sunshine upon bacterial life 
 in the great mass of Thames water should be nearly, if not quite, 
 imperceptible. 
 
 Upon the contrary, the influence of sunlight on the contents of 
 shallow reservoirs has been held to exert a very perceptible effect 
 on bacterial life in the water. Bacteriological cultivation is usu- 
 ally conducted in cupboards or incubators, from which the light is 
 
BACTERIAL CONTENTS OF VARIOUS WATERS. 55 
 
 rigorously excluded to avoid the inhibiting influence of light upon 
 the cultures, and this effect is known to be due to light indepen- 
 dent of heat from the sun's rays ; and one would have supposed 
 that in clear water, through which the light would penetrate for 
 some distance, a stronger influence would have been manifested 
 than is shown by the experiment quoted. 
 
 That the effect of sunlight is held by some experienced inves- 
 tigators to be of value in the reduction of bacterial life in polluted 
 waters, is apparent by the following quotation from a letter to the 
 author by Mr. Rud Schroder, inspector of the Hamburg Water- 
 Works: 
 
 11 In such places, where the winter temperatures do not vary from ours 
 [Hamburg], I believe open filters are to be preferred on account of their facil- 
 ity in being worked and cleaned ; and last, but not least, in regard to their bac- 
 teriological efficiency due to the rays of the sun." 
 
56 THE PURIFICATION OF WATER. 
 
 CHAPTER IV. 
 
 THE TYPHOID BACILLUS AND TYPHOID FEVER. 
 
 THE connection of b. typlwsus with the etiology of typhoid 
 fever is now so well established, and the relation of polluted water 
 supplies to typhoid fever so generally recognized, that a brief review 
 of the chief characteristics of this bacillus and allied organisms 
 found in water will not be inappropriate. 
 
 B. typhosus, obtained from a human spleen, will give the fol- 
 lowing characteristics : 
 
 (1) The bacillus will not liquefy gelatin. 
 
 (2) It will not coagulate sterilized milk. 
 
 (3) It will not produce gas when cultivated in glucose bouillon 
 in the fermentation tube. 
 
 (4) It will not give the indol reaction. 
 
 (5) When grown in a peptone solution containing potassium 
 nitrate, it is said to reduce nitrate to nitrite.* 
 
 (6) Under the microscope drop cultures show great activity 
 of the bacillus. (This the author finds depends upon the age of 
 the culture ; cultures near the original source (spleen) exhibit 
 greater motility than old cultures, and in degenerate cultures 
 motility seems to be no longer a property of the bacillus.) 
 
 (7) The vacuoles, or unstained spaces in the plasma, are rarely 
 absent. 
 
 (8) Filaments are often found in stained young cultures. 
 The invisible growth on sterilized potato is no longer regarded 
 
 as a test in differentiation of the typhoid bacillus. Other germs 
 will give the same effect, and the invisible growth is not constant 
 even with the typhoid organism. 
 
 In chemical composition the typhoid bacillus is not known to 
 differ from the harmless bacteria. Under the microscope it resem- 
 
 * Annual Report Massachusetts State Board of Health, 1891, p. 641. 
 
THE TYPHOID BACILLUS AND TYPHOID FEVER. 57 
 
 bles a diminutive rod with round ends, of length 'about three times 
 the width. The length, however, is not uniformly three times the 
 width ; for longer and shorter individuals will be seen on cover glass 
 preparations, and not infrequently long threads or undivided fila- 
 ments will be noticed in preparations from young cultures. Nu- 
 merous flagella, or whiplike appendages, spring from the cellulose 
 envelope, and endow the bacilius with motility. When stained with 
 carbol-fuchsin the bacillus appears as a brilliant red rod. 
 
 In looking for the typhoid bacillus in a sewage-polluted water, 
 in addition to certain non-pathogenic organisms resembling b. 
 typhosus in some respects, one is very liable to encounter two 
 other germs nearly identical with the typhoid bacillus. These are 
 b. coli communis and b. lactis aerogenes. 
 
 These three bacilli resemble (or differ from) each other in the 
 following respects : 
 
 (1) All are non-liquefiers. That is to say, all bacteria will 
 either (a), grow in gelatin and liquefy the material, or (), will 
 grow in the material without liquefying it. 
 
 (2) While there is a distinct difference between the three 
 bacilli in microscopic appearance, and b. coli communis and b. lactis 
 aerogenes seem to have definite proportions not easily mistaken, 
 the typhoid bacillus possesses such vagaries of proportion that at 
 times it seems to resemble closely the other germs. Although 
 b. coli communis is always thinner than the typhoid bacillus, and 
 b. lactis aerogenes always shorter, the length of b. coli communis 
 often agrees with the length (from a parallel culture) of the 
 typhoid bacillus, and the width of b. lactis aerogenes is about the 
 same as the typical typhoid rod, while all have rounded ends. 
 
 (3) All grow in gelatin stick cultures quite alike. ; b. lactis 
 aerogenes the quickest and most luxuriantly, and b. typhosus the 
 slowest and with least energy. Between b. typhosus and b. lactis 
 aerogenes, b. coli communis occupies a middle ground. All produce 
 a dirty white expansion on sloped agar. 
 
 (4) When cultivated in glucose bouillon in the fermentation 
 tube, these organisms grov/ equally well in the presence or absence 
 of air; the bouillon in the open leg (exposed through the cotton 
 plug to the air), and in the closed leg (from which all air has been 
 
58 THE PURIFICATION OF WATER. 
 
 removed during the sterilization of the contents of the tube), 
 exhibiting similar turbidity. When the allied germs, b. coli coin- 
 munis and b. lactis aerogenes, are cultivated in the fermentation 
 tube, an abundance of gas is produced in the closed leg, while the 
 typhoid bacillus, when so grown, produces no gas ; and this test 
 has been proposed by Dr. Theobald Smith * for the identification 
 of b. typhosus. 
 
 When the test is one to determine whether a certain germ is 
 the typhoid bacillus or b. coli communis, the fermentation tube will 
 settle the question ; but alone it will not determine whether a sus- 
 pected germ found in water is the typhoid bacillus or some other 
 organism. In Ohio River water the author has found bacilli which 
 grow in the fermentation tube quite like the typhoid bacillus, but 
 which are known by further tests not to be this germ. 
 
 (5) When the typhoid bacillus is grown in sterilized milk 
 having a slightly acid reaction, it will increase the acidity per- 
 ceptibly, but neither in the presence nor absence of heat will the 
 milk be coagulated ; while b. coli communis, when grown in milk, 
 produces a large increase of acidity, and sometimes coagulates the 
 milk in the tube at room temperature, and always coagulates the 
 milk upon the application of heat for a few minutes. 
 
 (6) When cultivated in gelatin or agar to which one or two 
 per cent of glucose is added, the typhoid bacillus will grow more 
 luxuriantly than in plain gelatin, but produces no gas, while b. coli 
 communis and b. lactis aerogenes, when so cultivated, produce an 
 abundance of gas. 
 
 (7) B. typhosus in drop cultures is possessed of great ac- 
 tivity, executing within the field of the microscope motions of 
 translation and rotation, and sinuous movements to and fro ; b. 
 coli communis has a sluggish motion, wholly unlike that of the 
 typhoid organism ; while b. lactis aerogenes is not possessed of 
 motility at all. 
 
 (8) All of these organisms are said to be found in the dejecta 
 of man, and b. coli communis and b. lactis aerogenes also in the 
 dejecta of animals. 
 
 * The Fermentation Tube, by Theobald Smith, Washington, B.C., 1893. 
 
THE TYPHOID BACILLUS AND TYPHOID FEVER. 
 
 59 
 
 The following experiments by the author wifh b. typhosus and 
 b. coli communis in sterilized milk throw some light upon the re- 
 spective acid-resisting properties of these organisms : 
 
 TESTS OF B. TYPHOSUS AND B. COLI COMMUNIS IN STERILIZED MILK. 
 
 (Acidity of 5 c. c. of milk tested before sterilization with a j 1 ^ normal solution of 
 caustic soda. All cultures grown at room temperature.) 
 
 Five cultures were used : 
 
 (A) B. typhosus obtained from Dr. T. M. Prudden, New York. 
 
 (B) B. typhosus obtained from Dr. O. L. Cameron, Cincinnati. 
 
 (C) B. typhosus obtained from Dr. O. L. Cameron, Cincinnati. 
 
 (D) B. coli communis obtained from Dr. O. L. Cameron, Cincinnati. 
 
 (E) B. typhosus obtained from spleen, Cincinnati Hospital. 
 
 DATE. 
 
 CULTURE. 
 
 ORIGINAL 
 ACIDITY OF 
 MILK. 
 
 DAYS 
 
 OF 
 
 GROWTH. 
 
 FINAL 
 ACIDITY OF 
 MILK.* 
 
 INCREASE 
 
 OF 
 
 ACIDITY.* 
 
 1894 
 
 
 
 
 
 
 Dec. 24, 
 
 A (2) 
 
 0.75 
 
 4| 
 
 0.775 
 
 0.025 
 
 " 
 
 B 
 
 0.75 
 
 4| 
 
 0.825 
 
 0.075 
 
 14 
 
 C 
 
 0.75 
 
 H 
 
 0.925 
 
 0.175 
 
 
 
 C 
 
 0.75 
 
 
 0.950 
 
 0.200 
 
 1895 
 
 
 
 
 
 
 Jan. 1, 
 
 A 
 
 0.80 
 
 4 
 
 0.910 
 
 0.110 
 
 " 
 
 A 
 
 0.80 
 
 4 
 
 0.930 
 
 0.130 
 
 
 
 B 
 
 0.80 
 
 4 
 
 1.100 
 
 0.300 
 
 Jan. 21, 
 
 B(2) 
 
 0.70 
 
 5 
 
 0.800 
 
 0.100 
 
 " 
 
 D 
 
 0.70 
 
 5 
 
 2.400 
 
 1.700 
 
 " 
 
 D 
 
 0.70 
 
 5 
 
 2.500 
 
 1.800 
 
 Jan. 26, 
 
 A 
 
 0.70 
 
 5 
 
 0.950 
 
 0.250 
 
 (C 
 
 A 
 
 0.70 
 
 5 
 
 0.850 
 
 0.150 
 
 Mar. 2, 
 
 E 
 
 0.70 
 
 6 
 
 0.900 
 
 0.200 
 
 
 
 E 
 
 0.70 
 
 6 
 
 0.880 
 
 0.180 
 
 Mar. 8, 
 
 A 
 
 0.70 
 
 4 
 
 0.975 
 
 0.275 
 
 " 
 
 E 
 
 0.70 
 
 4 
 
 0.925 
 
 0.225 
 
 June 6, 
 
 A 
 
 0.70 
 
 5 
 
 0.900 
 
 0.200 
 
 " 
 
 E(3) 
 
 0.70 
 
 5 
 
 1.000 
 
 0.300 
 
 While b. typJwsus will grow side by side with b. coli communis 
 in sterilized milk with a slight increase of the acidity of the milk, 
 the latter, on the contrary, will increase the acidity of milk to the 
 point of coagulation. 
 
 Culture " A " indicated an increase of 22 per cent in the acid- 
 
 * Acidity of milk stated in cubic centimeters of caustic soda solution. 
 
60 THE PURIFICATION OF WATER. 
 
 ity of milk ; culture " B " an increase of 21 per cent ; culture " C " 
 an increase of 25 per cent; culture "E" an increase of 30 per 
 cent ; and culture " D " (b. colt communis) indicated an increase 
 of 233 per cent in the acidity of milk. The increase of acidity of 
 sterilized milk by b. coli commimis as compared with b. typhosits is 
 nearly ten to one. 
 
 Aside from the able demonstration of its untenability by Dr. 
 Dunbar,* the theory which prevailed a few years since that an acid 
 solution might be made in which b. typhosus would develop, while 
 b. coli communis would perish, is clearly shown to be an impossi- 
 bility by the experiments detailed above. The difference in the 
 increase of acidity of sterilized milk by b. typhosus and b. coli com- 
 munis, however, is an important factor in the differentiation of 
 these organisms.f 
 
 A curious circumstance calculated to support Professor Lan- 
 kester's view of the origin of b. typhosus is, that while the colon 
 bacillus has been often found in polluted waters, the typhoid ba- 
 cillus has been found upon very rare occasions, and considerable 
 uncertainty surrounds quite a number of the alleged discoveries 
 of the latter bacillus in water supplies. Doubtless in all cases of 
 sewage pollution of water, the colon bacillus is much more numer- 
 ous than the typhoid bacillus, because the" former may come into 
 the sewage from many sources, while the latter can come only from 
 those suffering with typhoid fever, and even in such cases the 
 bacillus is said to be given off only during the earlier stages of 
 illness. 
 
 B. lactis aerogenes, likewise a bacillus of the intestine, by its 
 resemblance in several respects to b. typhosus and b. coli communis, 
 is very liable in the earlier stages of differentiation to be mistaken 
 for either. It, however, may be determined much easier that a 
 given organism is b. coli communis or b. lactis aerogenes, than that 
 it is b. typhosus ; and this difficulty in differentiating a germ sup- 
 posed to be b. typhosus, obtained outside the animal body, is a 
 
 * Zeitschrift fiir Hygiene, 1892, p. 485. ' 
 
 t Twenty-third Annual Report Massachusetts State Board of Health, p. 640. 
 
THE TYPHOID BACILLUS AND TYPHOID FEVER. 
 
 61 
 
 stumbling-block in the way of direct proof ttfat a given water 
 supply contains the typhoid germ. 
 
 The rapidity with which the bacillus usually acts when inocu- 
 lated into the lower animals renders these experiments an uncer- 
 tain test in differentiation, although the researches of Dr. Alessi * 
 with typhoid cultures on guinea pigs show that considerable time 
 may elapse between inoculation and death of these animals. These 
 experiments indicate that rats which die survived the infection from 
 12 to 36 hours, rabbits from 24 hours to 4 days, while guinea pigs 
 survived the inoculation from 8 hours to 13 days. 
 
 The investigations of Dr. Alessi show that while putrid gases, 
 i.e., sewer gas, may increase the susceptibility of the lower ani- 
 mals to typhoid fever infection, they cannot be considered as a 
 cause. Many animals were experimented upon ; and with the ex- 
 ception of rats first exposed to the influence of putrid gases and 
 then used as controls, all the animals not inoculated with the 
 typhoid culture recovered. The following table contains a rfeumt 
 of the mortality of the animals from these experiments : 
 
 EXPERIMENTS WITH CULTURES OF B. TYPHOSUS AND B. COLI COMMUNIS. 
 
 
 
 PERCENTAGE OF MORTALITY. 
 
 CULTURE USED. 
 
 ANIMALS. 
 
 Inoculated 
 
 Control 
 
 
 
 Animals. 
 
 Animals. 
 
 B. typhosus (A), 
 
 Rats, 
 
 75 
 
 7 
 
 " 
 
 Guinea Pigs, 
 
 79 
 
 
 
 
 
 Rabbits, 
 
 100 
 
 
 
 B. typhosus (B), 
 
 Guinea Pigs, 
 
 80 
 
 
 
 < 
 
 Rabbits, 
 
 70 
 
 
 
 B. coli communis, 
 
 Guinea Pigs, 
 
 83.3 
 
 
 
 It is held at the present time by the best-informed students 
 along this line, that if a given source of water supply is exposed to 
 constant or even occasional sewage contamination, and that such 
 sewage contains the dejecta from typhoid patients in hospitals and 
 residences, it is not necessary to have direct proof of the presence 
 of the typhoid bacillus in such water to justify its condemnation. 
 
 * "On Putrid Gases as Predisposing Causes of Typhoid Infection," by Dr. Giuseppe 
 Alessi, Journal of the Sanitary Institute, London, January, 1896, p. 505. 
 
62 THE PURIFICATION OF WATER. 
 
 The circumstantial evidence that the organism 'is in such water 
 is sufficient, and the failure to find the bacillus upon test should 
 not be taken as evidence of its non-existence. 
 
 The typhoid bacillus, originally isolated and described by 
 Eberth, has been made the subject of careful study by Koch, 
 Gaffky, Frankell, and Simmonds, and latterly by Joseph Sanarelli, 
 of the Pasteur Institute, Paris. Several papers by Sanarelli have 
 appeared in discussion of the organism, its connection with the 
 etiology of typhoid fever, the symptoms and lesions produced, and 
 results of inoculation in the lower animals, of which the more 
 important deductions and opinions are worthy of mention. 
 
 The so-called typhotoxin, described by Brieger as a product of 
 the vital activity of the Eberth bacillus, is considered by Sanarelli 
 only as an ordinary product of decomposition, arising from the 
 changes which occur in the albuminoid substances of the culture 
 media, or to bacterial poisons previously existing in the culture. 
 
 Of the Stas-Otto method by which typhotoxin is obtained it 
 has been said : * 
 
 " However, the method is not free from criticism. The great number of 
 chemical manipulations to which the organic matter is subjected is liable to 
 lead to the formation of some basic substances, and to the destruction of 
 others. One is justified in considering the isolated base as preexisting in the 
 original material, only when it produces symptoms identical with those caused 
 by the substance from which it is extracted." 
 
 Vaughan and Novy, however, assume that Brieger has actually 
 isolated the poisonous product of the typhoid bacillus ; while Sana- 
 relli rejects typhotoxin, so-called, as a product of the Eberth germ. 
 Sanarelli calls attention to the fact that " recent investigators have 
 shown that evaporation of the albuminous liquids in the presence 
 of hydrochloric acid and their subsequent extraction with alcohol 
 is alone sufficient to produce bodies considered by Brieger as pto- 
 maines, and the typhotoxin does not produce a morbid state com- 
 parable with that of typhoid fever," and concludes that all attempts 
 to ascertain the chemical nature of the poison produced by the 
 typhoid bacillus are failures. 
 
 * Ptomaines and Leucomaines, by Vaughan and Novy, Philadelphia, 1891, p. 170. 
 
THE TYPHOID BACILLUS AND TYPHOID FEVER. 63 
 
 Sanarelli argues that no matter by what channel the typhoid 
 bacillus enters the system, its seat of operation is not the small 
 intestine (Peyers glands), as heretofore supposed, but the spleen. 
 Here it grows and elaborates the toxin which is taken into the cir- 
 culation, and produces certain local effects which are characteristic 
 of typhoid fever. He holds that the toxic product of the growth 
 of b. typJwsus in the spleen, when taken into the circulation, para- 
 lyzes the walls of the intestine, and destroys its powers of resistance 
 to the action of b. coli communis, and that all local effect in the 
 ileum is to be charged to the latter organism, and not, as has been 
 generally supposed, to the action of b. typhosus on the mucosa and 
 vessels of the intestine. 
 
 He maintains that all local symptoms of typhoid fever are like 
 the symptoms in the ileum, altogether due to the toxic properties 
 imparted to the circulation by the growth and development of b. 
 typhosus in the spleen. He claims that the typhoid bacillus has 
 never been isolated from the dejecta, nor from the anatomical 
 changes in the intestine, and argues that if typhoid fever has its 
 beginning and end in the digestive tract, why do we not find the 
 b. typhosus there from the very beginning, before the symptoms 
 characteristic of the disease are noticed. The diarrhea, he insists, 
 is maintained and aggravated by b. coli commnnis. This remark- 
 able series of papers contains the following conclusions : 
 
 " The extraordinary multiplication of the colon bacillus, and its tendency 
 to destroy all other bacteria and become the sole representative of the intesti- 
 nal species, are the results of an active biologic work, incessant and complex. 
 . . . When the poison (elaborated by b. typhosus} has attained the extreme 
 limit of its tolerance by the subject, the fever ceases, and collapse ends the 
 scene. It is this period of collapse which we reproduce experimentally on 
 animals. In them the typhoid virus permeates the system, and manifests its 
 effects too rapidly to give it time to react by means of fever during the early 
 stages of intoxication. If the bacillus of Eberth could produce its toxin in the 
 human system as rapidly as the cholera spirilla do theirs, typhoid fever would 
 be like cholera, a disease short of duration and without fever." 
 
 According to Sanarelli, by reason of their lower powers of 
 resistance, the action of b. typhosus on the lower animals is too 
 rapid to produce the train of effects which are recorded of this 
 disease in man. 
 
64 THE PURIFICATION OF WATER. 
 
 To the medical practitioner these conclusions of Sanarelli, if 
 confirmed by further investigation, should possess great value, 
 and be an aid to him in revising his methods of treatment of 
 typhoid fever. To the sanitarian they also possess interest, as 
 showing the manner in which a certain bacterium (b. coli corn- 
 munis), always in the human intestine, and frequently found in 
 sewage -polluted waters, may indirectly be endowed with extra 
 pathogenic powers, and sustain a relationship to typhoid fever 
 not heretofore suspected. 
 
 The typical typhoid bacillus, the bacillus of Eberth, is found in 
 the spleen, and occasionally in some other organs of one who has 
 died during the early stages of typhoid fever. This organism, as 
 is well known, has morphological and biological characteristics 
 unlike the colon bacillus ; and considering the well-founded doubts 
 of the several alleged discoveries of the true typhoid germ in water 
 supplies, while many investigators have isolated the colon bacillus 
 from this source, some strength is imparted to Professor Lankes- 
 ter's assumption that b. typhosus may be an exacerbated form of 
 b. coli communis. 
 
 It has been stated that Malvoz of Liege, by successive culti- 
 vations of b. coli commtmis in a slightly acid bouillon, produced a 
 species showing the characteristics of Eberth's bacillus. 
 
 The influence of environment on species is well known ; and may 
 it be possible that b. coli communis, when taken into the human 
 system through the medium of drinking-water, in certain persons 
 finds there the conditions favorable to its development into what 
 we know as b. typhosus ? Certain schools * refuse to recognize a 
 difference between these two organisms, although they do not go 
 so far as Professor Lankester, and assume that one may become 
 the other. They regard the two bacilli as different forms of the 
 same species, and treat b. coli communis as a cause of typhoid fever. 
 
 Dr. Jordan,! of the University of Chicago, in a very able paper 
 on the characteristics of the typhoid bacillus, says : - 
 
 * Sanarelli : Annales de r Institute Pasteur, Nov. 25, 1892. 
 
 t The Identification of the Typhoid Fever Bacillus, by Edwin O. Jordan, Ph.D., Chicago, 
 
THE TYPHOID BACILLUS AND TYPHOID FEVER. 65 
 
 " While the close similarity of the colon bacillus anft the typhoid bacillus 
 is necessarily recognized 'by every one, and while it is admitted that there is no 
 single criterion that absolutely distinguishes the latter from all the perplexing 
 'varieties' and 'related forms,' it is nevertheless maintained by many that the 
 sum total of the morphological and physiological characters presents a true 
 and unmistakable picture of the specific organism of typhoid fever. These 
 investigators hold that, although the varieties may approach more or less 
 closely to the typical typhoid germ, they may always be distinguished from it 
 by at least some one character which is not shared by the genuine typhoid 
 organism." 
 
 In his summary on the existing information upon the differen- 
 tiation of the typhoid organism from the colon bacillus and allied 
 germs, among other conclusions he states : 
 
 (1) " There is usually found in the spleen and other organs of an indi- 
 vidual dying with typhoid fever, a bacillus which possesses certain definite 
 morphological and physiological characters. . . . 
 
 (5) "The cases of alleged conversion of one 'species' or 'variety' (b. coli 
 communis} into another (b. typhosus} do not carry conviction, and are suscept- 
 ible of other interpretations than those advanced regarding them." 
 
 So far as protection to our public water supplies is concerned, 
 this diversity of opinion on the typhoid bacillus as an independ- 
 ent organism can have little weight. The means adopted for the 
 exclusion of b. typhosus from our drinking-waters will also exclude 
 the colon bacillus. At least, it is not now known that a filter 
 properly constructed and operated will not be equally effective in 
 restraining the passage of either germ ; and methods of sterilization 
 such as are in use on sea-going vessels are bound to eliminate both 
 organisms from our drinking-waters. There is, however, this dif- 
 ference to be considered between the generally accepted theory 
 of the transmission of the typhoid bacillus from the sick to the 
 well through the direct sewage pollution of drinking-waters, and 
 the theory of Professor Lankester of the colon bacillus under 
 proper conditions becoming the typhoid bacillus, that if the latter 
 should be substantiated, the exposure to typhoid fever from water 
 sources is much greater than has generally been supposed ; and no 
 surface water supply, the drainage area of which is inhabited by 
 domestic animals, can be regarded as proof against a possible con- 
 tamination by b. coli communis. 
 
66 
 
 THE PURIFICATION OF WATER. 
 
 The percentage of mortality from typhoid fever is variously 
 stated in text-books, and some statistics from the later experience 
 of American cities with this disease are given in the table which 
 follows : 
 
 TYPHOID FEVER MORTALITY. 
 
 CITY. 
 
 PERIOD TAKEN. 
 
 CASES 
 REPORTED. 
 
 DEATHS. 
 
 PERCENTAGE 
 
 OF 
 
 MORTALITY. 
 
 Lowell, Mass., 
 
 Sept., 1890, to Jan. 1891, 
 
 550 
 
 89 
 
 16.2 
 
 
 
 Nov., 1892, to Feb., 1893, 
 
 141 
 
 30 
 
 21.3 
 
 Lawrence, Mass., 
 
 Nov., 1892, to Feb., 1893, 
 
 141 
 
 28 
 
 20.0 
 
 Springfield, Mass., 
 
 July, 1892, to Sept., 1892, 
 
 155 
 
 32 
 
 20.6 
 
 St. Louis, Mo., 
 
 April, 1892, to Mar., 1893, 
 
 3,624 
 
 514 
 
 14.2 
 
 < ( 
 
 Aug., 1892, to Jan., 1893, 
 
 3,455 
 
 453 
 
 13.1 
 
 Pittsburg, Pa., 
 
 1891, 
 
 1,047 
 
 248 
 
 23.7 
 
 
 
 1892, 
 
 1,145 
 
 256 
 
 22.4 
 
 <( 
 
 1893, 
 
 2,398 
 
 294 
 
 12.3 
 
 Schenectady, N.Y., 
 
 July, 1890, to April, 1891, 
 
 300 
 
 70 
 
 23.3 
 
 Cincinnati Hospital, 
 
 1891, 
 
 85 
 
 11 
 
 13.0 
 
 
 
 1892, 
 
 59 
 
 7 
 
 12.0 
 
 <t 
 
 1893, 
 
 108 
 
 4 
 
 3.7 
 
 
 
 1894, 
 
 117 
 
 14 
 
 12.0 
 
 
 
 1895, 
 
 98 
 
 10 
 
 10.2 
 
 It is a fact well known to physicians and investigators, that 
 typhoid fever is everywhere a disease which sets in late in sum- 
 mer or early in autumn, and continues above the normal rate 
 until midwinter, after which it usually shows a decline, falling to 
 the lowest case and death rates during the months of spring. The 
 theory of Pettenkofer and others of the Munich school,* that the 
 disease rises and falls in intensity as the level of the ground water 
 falls and rises, is partly supported by the usual seasonal distribu- 
 tion of typhoid. 
 
 In spring the level of ground water is high, and the typhoid 
 rates low. In the autumn the level of the ground water is low, 
 while the typhoid rates are high ; but so many proofs are at hand 
 showing high typhoid rates with high levels of ground water, and 
 low typhoid rates accompanied by low levels of ground water, that 
 the Pettenkofer theory of a connection between the level of the 
 
 * Cholera, by Dr. Max Von Pettenkofer, translated by Dr. T. M. Hime, London, 1893 
 (chart facing title-page). 
 
THE TYPHOID BACILLUS AND TYPHOID FEVER. 67 
 
 ground water and typhoid fever is not of general application, and 
 cannot point to a probable general cause for this disease in the 
 lowering of the level of ground water. 
 
 Another theory of the cause of the rise in the typhoid fever 
 rates during autumn is based on the assumption that the early 
 rains carry into the streams and lakes which constitute so many 
 sources of water supply, sewage and offal which has accumulated 
 upon the ground and along the banks of streams during the period 
 of drought. Opposed to this theory, however, is the almost uni- 
 form condition that the typhoid rates begin to rise before the fall 
 rains occur. Indeed, the maximum rate is often reached in the 
 autumn, before rains have fallen sufficient in volume to swell the 
 streams, and wash organic refuse from the banks. 
 
 Dr. Woodhead * has indirectly suggested a theory which, prop- 
 erly expanded, seems to furnish a good reason for the seasonal 
 influence on the typhoid rate ; viz., that the development of the 
 typhoid bacillus in water will depend upon the temperature of 
 the water in which it is migrating, low temperature discoura- 
 ging the growth of the bacillus, while high temperature favors it. 
 During the early autumn the lakes and rivers are at the highest 
 temperature, which continues with slow diminution (in the tem- 
 perate zone) well into the winter. During this period of high 
 natural water temperature, typhoid fever usually rises to its great- 
 est intensity for the year ; and when the temperature of the water 
 naturally declines, and approaches its lowest point after midwinter, 
 the typhoid rates also subside and reach their minimum. 
 
 This theory of the rise and fall of typhoid rates does not depend 
 upon floods nor upon the level of the ground water, and is entirely 
 consistent with the information upon the growth of the typhoid 
 bacillus in the human body and in the usual culture media. 
 
 The influence of proper preventive measures on typhoid fever 
 is shown by the experience of Vienna and Munich. 
 
 Professor Von Zeimssen f of the latter city a few years ago 
 stated " that the reduction in the cases in the hospital had almost 
 
 * Royal Commission on Metropolitan Water Supply, London, 1893, Minutes of Evi- 
 dence, p. 506. 
 
 t On Typhoid Fever in Baltimore, by Dr. William Osier, 1892. 
 
68 THE PURIFICATION OF WATER. 
 
 changed the character of the service, and they had scarcely patients 
 enough to illustrate the disease in the clinical courses." Dr. Osier 
 is disposed to credit this reduction in the typhoid rates in these 
 cities to improved sewerage, when it is really due largely to the 
 remarkable improvements in their public water supplies, Munich 
 having abandoned its old sources in wells and the River Isar for 
 mountain springs in the Mangfall valley, and Vienna having aban- 
 doned its wells and the Danube for the Schneeberg springs in the 
 Alps. 
 
 Improvements in sewerage should reduce the susceptibility of a 
 people to typhoid infection, but the cause is removed when pure 
 water is substituted for polluted water. 
 
 The channels of typhoid fever infection for a large city are 
 clearly indicated in a recent paper by Dr. J. J. Reincke, On the 
 Epidemiology of Typhoid Fever in Hamburg and Altona ; * and as 
 applicable especially to Hamburg before filtration of the Elbe water 
 was adopted, he mentions : 
 
 (1) Importation of typhoid by way of the sea and the upper Elbe, and 
 infection of the water of the river in front of the city by travelers. 
 
 (2) Infection of the water of the Elbe by patients in the city, whose dis- 
 infected dejecta is carried by the sewage into the river. 
 
 (3) Infection of the Elbe water from people on the ships, city dock labor- 
 ers, and bathers. 
 
 (4) Infection by means of food which had come in contact with the Elbe 
 water. 
 
 (5) Infection through the unfiltered Elbe water distributed to the people 
 prior to May, 1893. 
 
 (6) Infection by patients and food brought into Hamburg from neighbor- 
 ing places. 
 
 (7) Secondary infection within the city by direct transmission, or infection 
 of food and wells. [Causes which he holds were peculiarly numerous and effec- 
 tive in the second half of the great epidemics.] 
 
 Concerning the seasonal distribution of typhoid, he shows how 
 the number of cases increase in autumn and decline towards spring. 
 He points to the fact that in Germany the years 1857, 1865, and 
 the early 70's were years of high typhoid fever rates, and the years 
 of 1860 and 1867 were years of phenomenally low typhoid fever 
 
 * Hamburg, 1896. 
 
THE TYPHOID BACILLUS AND TYPHOID FEVER. 69 
 
 rates ; and suggests that " there must be certain 'influences at work 
 (here) that are more far-reaching than the predisposing causes." 
 
 He mentions conditions heretofore remarked, that the epi- 
 demics of typhoid which originate in the autumn have a gradual 
 decline, while epidemics which originate during the late winter or 
 early spring months show a sharp rise and fall of intensity and a 
 sudden disappearance. Seasons of drought with low levels of ground 
 water he considers as favoring high typhoid rates, while wet sea- 
 sons and high ground water levels he regards as unfavorable. Dur- 
 ing dry seasons there is greater body heat with more thirst, and a 
 larger consumpion of water for drinking and bathing purposes, 
 which increases the liability to infection. 
 
 Dr. Reincke combats Pettenkofer's theory of the cause of 
 typhoid fever, while agreeing with him that the proper prophy- 
 laxis for the disease is to be found in high quality of public water 
 supplies and efficient sewerage and drainage. The very signifi- 
 cant statement is made, that if typhoid fever epidemics are pre- 
 vented in the cities there will be none in the country, and as a 
 consequence there can be no reimportation of cases into the cities. 
 
 The author, after careful investigation, reached the conclusion 
 several years ago that the sewage-polluted waters of the cities were 
 largely responsible for typhoid fever in the country districts, and 
 agrees with Dr. Reincke that the typhoid rates of any city is a 
 measure of the efficiency of its works of sanitation. 
 
 Speaking of Hamburg, he says "that the present favorable 
 and heretofore unattainable status of typhoid fever is to be largely 
 credited to the filtered water supply, as indicated by the fact that 
 this disease has not diminished among the people on the vessels 
 arriving at the port." 
 
70 THE PURIFICATION OF WATER. 
 
 CHAPTER V. 
 
 CLASSIFICATION OF CITIES BY TYPHOID FEVER 
 STATISTICS. 
 
 IT has been -held (and correctly in the author's opinion) that 
 the best test of the quality of a city water supply was the typhoid 
 fever rates of that city. Thus a city with water of high quality 
 should show low typhoid rates, and a city with water of known 
 sewage pollution should show high typhoid rates. The final test 
 of all public water supplies is the influence of these on the health 
 of the consumers ; and in order to institute a comparison of cities 
 upon the basis of quality of water supply, the author proposed, in a 
 lecture recently delivered before the faculty and students of the 
 University of Illinois,* to classify the larger cities of the world, 
 embraced within the scope of modern health statistics, upon their 
 typhoid fever death rates. 
 
 Thus cities of the first class must show a death rate from 
 typhoid fever of not more than 10 per 100,000 of population 
 living. 
 
 Cities of the second class must show a rate not higher than 20 
 per 100,000 of population living. 
 
 Cities of the third class must show a rate not higher than 30 
 per 100,000 of population living, and in like manner by tens until 
 the sixth class was reached. All cities having a typhoid death rate 
 in excess of 60 per 100,000 of population living are grouped in 
 the seventh class. The classification is made upon the average 
 rates for the years 1890 to 1896, inclusive, for the principal cities 
 of the United States, Canada, and Europe, including two cities in 
 Egypt and two cities in Australia. The statistics from which the 
 classification is made will be found in the table of Typhoid Fever 
 Statistics, Appendix A. 
 
 * Water Supply of Cities, Champaign, 111., 189G. 
 
CLASSIFICA riON OF CITIES B Y TYPHOID FE VER STA TISTICS. 71 
 
 CLASS I. CONTAINING ALL CITIES SHOWING A TYPHOID FEVER DEATH 
 RATE OF 1O OR LESS PER 1OO.OOO OF POPULATION LIVING. 
 
 CLASS I. 
 
 CITY. 
 
 SOURCE OF WATER SUPPLY. 
 
 POPULATION. 
 
 TYPHOID 
 FEVER 
 DEATH 
 RATE. 
 
 Hague, 
 
 From sand dunes, 
 
 187,545 
 
 4.7 
 
 Rotterdam, 
 
 Filtered water from River Maas, 
 
 276.338 
 
 5.7 
 
 Munich, 
 
 Spring water from Mangfall Valley, 
 
 406,000 
 
 6.0 
 
 Dresden, 
 
 Filter wells and gallery by River Elbe, 
 
 342,340 
 
 6.2 
 
 Vienna, 
 
 Springs in the Schneeberg, 
 
 1,526,623 
 
 6.6 
 
 Berlin, 
 
 Filtered water from River Spree and 
 
 
 
 
 Lake Tegel, 
 
 1,695,313 
 
 7.1 
 
 Copenhagen, 
 
 Filtered from wells and springs, 
 
 333,714 
 
 9.0 
 
 CLASS II. 
 
 CITY. 
 
 SOURCE OF WATER SUPPLY. 
 
 POPULATION. 
 
 TYPHOID 
 FEVER 
 DEATH 
 RATE. 
 
 Christiania, 
 
 
 182,856 
 
 10.6 
 
 Breslau, 
 
 Filtered from River Oder, 
 
 377,062 
 
 10.7 
 
 Amsterdam, 
 
 Haarlem dunes, and River Vecht, filtered, 
 
 489,496 
 
 11.9 
 
 Stockholm, 
 
 Lake and well water, 
 
 267,100 
 
 12.3 
 
 London, 
 
 Kent wells, and filtered water from Riv- 
 
 
 
 
 ers Thames and Lea, 
 
 4,421,955 
 
 14.4 
 
 Trieste, 
 
 
 161,866 
 
 14.7 
 
 Edinburgh, 
 
 Impounded and filtered water from Pent- 
 
 
 
 
 land Hills, 
 
 276,514 
 
 16.4 
 
 Hamburg, 
 
 Filtered water from River Elbe, 
 
 625,552 
 
 17.7 
 
 Brooklyn, 
 
 Impounded water, and open and driven 
 
 
 
 
 wells, 
 
 1,140,000 
 
 18.0 
 
 New York, 
 
 Impounded water from Croton and Bronx 
 
 
 
 
 Rivers, 
 
 1,934,077 
 
 19.3 
 
 CLASS III. 
 
 
 
 
 TYPHOID 
 
 
 
 
 FEVER 
 
 CITY. 
 
 SOURCE OF WATER SUPPLY. 
 
 POPULATION. 
 
 DEATH 
 
 
 
 
 RATE. 
 
 Paris, 
 
 Rivers Seine, Marne, and Vanne, Ourcq 
 
 
 
 
 Canal, artesian wells, and springs, 
 
 2,511,629 
 
 22.0 
 
 Sydney, N.S.W. 
 
 Impounded water from Upper Nepean 
 
 
 
 
 River, 
 
 423,600 
 
 22.0 
 
 Glasgow, 
 
 Loch Katrine, 
 
 705,052 
 
 23.0 
 
 Buda-Pest, 
 
 Ground water from wells, 
 
 579,275 
 
 23.0 
 
72 
 
 THE PURIFICATION OF WATER. 
 
 CL AS S III. Continued. 
 
 CITY. 
 
 SOURCE OF WATER SUPPLY. 
 
 POPULATION. 
 
 TYPHOID 
 FEVER 
 DEATH 
 RATE. 
 
 Brussels, 
 
 
 518,387 
 
 23.5 
 
 Manchester, 
 
 Lake Thirlmere, 
 
 530,000 
 
 25.7 
 
 New Orleans, 
 
 Drinking-water from tanks and cisterns, 
 
 275,000 
 
 25.9 
 
 Altona, 
 
 Filtered from River Elbe, 
 
 148,934 
 
 26.8 
 
 Davenport, 
 
 Mechanical filter, Mississippi River, 
 
 35,000 
 
 28.6 
 
 Venice, 
 
 Impounded water, 
 
 163,254 
 
 28.7 
 
 Milwaukee, 
 
 Lake Michigan, 
 
 257,500 
 
 29.3 
 
 Brisbane, Qu. 
 
 
 93,657 
 
 29.7 
 
 CLASS IV. 
 
 DATE. 
 
 SOURCE OF WATER SUPPLY. 
 
 POPULATION. 
 
 TYPHOID 
 FEVER 
 DEATH 
 RATE. 
 
 Detroit, 
 
 Detroit River, 
 
 279,000 
 
 30.1 
 
 Boston, 
 
 Lake Cochituate and Sudbury River, 
 
 508,694 
 
 32.6 
 
 Buffalo, ' 
 
 Niagara River, at head, 
 
 350,000 
 
 34.3 
 
 Turin, 
 
 
 344,203 
 
 34.3 
 
 Rome, 
 
 Fontanadi Trevi, Aqua Felice, and Paoli, 
 
 473,296 
 
 35.7 
 
 Dayton, 
 
 Driven wells, 
 
 85,000 
 
 36.0 
 
 Liverpool, 
 
 Lake Vyrnwy (Wales), 
 
 632,000 
 
 36.3 
 
 Providence, 
 
 Pawtuxet River, 
 
 150,000 
 
 36.4 
 
 Covington, 
 
 Ohio River, 
 
 50,000 
 
 36.6 
 
 Newark, 
 
 Impounded water from Pequannock 
 
 
 
 
 River, 
 
 230,000 
 
 38.1 
 
 San Francisco, 
 
 Impounded water from mountain springs, 
 
 330,000 
 
 38.4 v, 
 
 St. Louis, 
 
 Mississippi River, 
 
 570,000 
 
 39.0 
 
 CLASS V. 
 
 CITY. 
 
 SOURCE OF WATER SUPPLY. 
 
 POPULATION. 
 
 TYPHOID 
 FEVER 
 DEATH 
 RATE. 
 
 Prague, 
 
 
 364,632 
 
 41.4 
 
 Baltimore, 
 
 Gunpowder River, Lake Roland, 
 
 507,398 
 
 43.6 
 
 Nashville, 
 
 Filter gallery, Cumberland River, 
 
 87,754 
 
 44.7 
 
 Philadelphia, 
 
 Schuylkill and Delaware Rivers, 
 
 1,188,793 
 
 45.0 
 
 Cleveland, 
 
 Lake Erie, 
 
 330,279 
 
 46.4 
 
 Denver, 
 
 South Platte River and Marston Lake, 
 
 150,000 
 
 47.2 
 
 Toronto, 
 
 Lake Ontario, 
 
 196,666 
 
 49.3 
 
 Cincinnati, 
 
 Ohio River, 
 
 341,000 
 
 49.4 
 
CLASSIFICA TION OF CITIES B Y TYPHOID FE VER STA TISTICS. 73 
 
 CLASS VI. 
 
 CITY. 
 
 SOURCE OF WATER SUPPLY. 
 
 POPULATION. 
 
 TYPHOID 
 FEVER 
 DEATH 
 RATE. 
 
 Dublin, 
 
 River Vartry, impounded water, filtered, 
 
 349,594 
 
 52.3 
 
 Quincy, 
 
 Mechanical filters, Mississippi River, 
 
 42,000 
 
 53.6 
 
 Moscow, 
 
 Mytisch springs and ponds, Moscow and 
 
 
 
 
 Yanza Rivers, 
 
 753,469 
 
 54.4 
 
 Newport, 
 
 Ohio River, 
 
 30,000 
 
 57.7 
 
 Knoxville 
 
 Mechanical filters, Tennessee River, 
 
 45,000 
 
 59.7 
 
 (with suburbs), 
 
 
 
 
 Milan, 
 
 
 441,948 
 
 59.7 
 
 CLASS VII. 
 
 CITY. 
 
 SOURCE OF WATER SUPPLY. 
 
 POPULATION. 
 
 TYPHOID 
 FEVER 
 DEATH 
 RATE. 
 
 Indianapolis, 
 
 Driven wells, and White River, 
 
 165,000 
 
 64.5 
 
 Chattanooga, 
 
 Mechanical filters, Tennessee River, 
 
 . 40,000 
 
 68.2 
 
 Washington, 
 
 Potomac River, 
 
 278,150 
 
 71.1 
 
 Chicago, 
 
 Lake Michigan, 
 
 1,619,226 
 
 71.2 
 
 Jersey City, 
 
 Passaic and Pequannock Rivers, 
 
 187,098 
 
 72.5 
 
 Louisville, 
 
 Ohio River, 
 
 211,100 
 
 74.3 
 
 Lawrence, Mass., 
 
 Filtered water fromMerrimac River, 
 
 55,000 
 
 75.3 
 
 St. Petersburg, 
 
 Filtered water from River Neva, 
 
 954,400 
 
 77.2 
 
 Lowell, Mass., 
 
 Merrimac River, driven wells, 
 
 85,700 
 
 77.6 
 
 Pittsburgh, 
 
 Alleghany River, 
 
 280,000 
 
 84.2 
 
 Atlanta, 
 
 Mechanical filters, Chattanooga 
 
 
 
 
 River, 
 
 110,000 
 
 85.1 
 
 Alexandria, Egypt, 
 
 River Nile, by canal, 
 
 231,396 
 
 143.4 
 
 Cairo, Egypt, 
 
 River Nile, by canal, 
 
 374,838 
 
 168.3 
 
 CLASSIFICATION BASED ON LAST YEAR REPORTED (1896). 
 CLASS I. 
 
 CITY. 
 
 POPULATION. 
 
 RATE. 
 
 CITY. 
 
 POPULATION. 
 
 RATE. 
 
 Amsterdam, 
 
 489,496 
 
 3 
 
 Vienna, 
 
 1,526,623 
 
 5 
 
 Munich, 
 
 406,000 
 
 3 
 
 Hamburg, 
 
 625,552 
 
 6 
 
 Hague, 
 
 187,545 
 
 4 
 
 Stockholm, 
 
 267,100 
 
 6 
 
 Dresden, 
 
 342,340 
 
 4 
 
 Copenhagen, 
 
 333,714 
 
 7 
 
 Berlin, 
 
 1,695,313 
 
 5 
 
 Breslau, 
 
 377,062 
 
 8 
 
74 
 
 THE PURIFICATION OF WATER. 
 
 CLASS II. 
 
 CITY. 
 
 POPULATION. 
 
 RATE. 
 
 CITY. 
 
 POPULATION. 
 
 RATE. 
 
 Paris, 
 
 2,511,629 
 
 11 
 
 Lawrence, 
 
 55,000 
 
 15 
 
 Rotterdam, 
 
 276,338 
 
 12 
 
 Edinburgh, 
 
 276,514 
 
 16 
 
 Altona, 
 
 148,934 
 
 13 
 
 New York, 
 
 1,934,077 
 
 16 
 
 Trieste, 
 
 161,886 
 
 13 
 
 Milwaukee, 
 
 257,500 
 
 18 
 
 London, 
 
 4,421,955 
 
 14 
 
 Brussels, 
 
 518,387 
 
 18 
 
 Brooklyn, 
 
 1,140,000 
 
 15 
 
 St. Louis, 
 
 570,000 
 
 19 
 
 CLASS III. 
 
 CITY. 
 
 POPULATION. 
 
 RATE. 
 
 CITY. 
 
 POPULATION. 
 
 RATE. 
 
 Buffalo, 
 
 350,000 
 
 20 
 
 Dayton, 
 
 85,000 
 
 25 
 
 Detroit, 
 
 279,000 
 
 20 
 
 Quincy, 
 
 42,000 
 
 26 
 
 Davenport, 
 
 35,000 
 
 20 
 
 Venice, 
 
 163,254 
 
 27 
 
 Sydney, 
 
 423,600 
 
 20 
 
 Providence, 
 
 150,000 
 
 27 
 
 Newark, N.J., 
 
 230,000 
 
 21 
 
 Rome, 
 
 473,296 
 
 27 
 
 Glasgow, 
 
 705,052 
 
 23 
 
 Prague, 
 
 364,632 
 
 28 
 
 Manchester, 
 
 530,000 
 
 23 
 
 Toronto, 
 
 196,666 
 
 28.5 
 
 Turin,- 
 
 344,203 
 
 24 
 
 Buda-Pest, 
 
 579,275 
 
 29 
 
 CLASS IV. 
 
 CITY. 
 
 POPULATION. 
 
 RATE. 
 
 CITY. 
 
 POPULATION. 
 
 RATE. 
 
 Chattanooga, 
 
 40,000 
 
 30 
 
 Liverpool, 
 
 632,000 
 
 32 
 
 San Francisco, 
 
 330,000 
 
 31 
 
 Christiania, 
 
 182,856 
 
 33 
 
 Boston, 
 
 508,694 
 
 32 
 
 New Orleans, 
 
 275,000 
 
 33 
 
 Covington, 
 
 50,000 
 
 32 
 
 Philadelphia, 
 
 1,188,793 
 
 34 
 
 Knoxville, 
 
 37,000 
 
 32 
 
 Baltimore, 
 
 507,398 
 
 37 
 
 (without suburbs), 
 
 
 
 
 
 
 CLASS V. 
 
 CITY. 
 
 POPULATION. 
 
 RATE. 
 
 CITY. 
 
 POPULATION. 
 
 RATE. 
 
 Indianapolis, 
 
 165,000 
 
 41 
 
 Dublin, 
 
 349,594 
 
 45 
 
 Lowell, Mass., 
 
 85,700 
 
 42 
 
 Chicago, 
 
 1,619,226 
 
 46 
 
 Cleveland, 
 
 330,279 
 
 43 
 
 Moscow, 
 
 753,469 
 
 46 
 
 Louisville, 
 
 211,100 
 
 45 
 
 Cincinnati, 
 
 341,000 
 
 48 
 
 CLASS VI. 
 
 CITY. 
 
 POPULATION. 
 
 RATE. 
 
 Washington, 
 Nashville, 
 Milan, 
 
 278,150 
 87,754 
 441,948 
 
 51 
 55 
 55 
 
CLASSIFICA TION OF CITIES B Y TYPHOID FE VER STA TISTICS. 75 
 
 CLASS VII. 
 
 CITY. 
 
 POPULATION. 
 
 RATE. 
 
 CITY. 
 
 POPULATION. 
 
 RATE. 
 
 Atlanta, 
 
 110,000 
 
 60 
 
 Newport, Ky., 
 
 30,000 
 
 63.0 
 
 Brisbane, 
 
 93,657 
 
 60 
 
 Alexandria, 
 
 231,396 
 
 89 
 
 Pittsburg, 
 
 280,000 
 
 61 
 
 Cairo, 
 
 374,838 
 
 141 
 
 Denver, 
 
 150,000 
 
 61 
 
 St. Petersburg, 
 
 954,400 
 
 142 
 
 Jersey City, 
 
 187,098 
 
 61.5 
 
 
 
 
 A review of the statistics furnished by the tables reveals some 
 interesting facts : 
 
 No city in the United States appears in the first class, and only 
 two cities (Brooklyn and New York) appear in the second class. 
 All other cities in these two classes are found abroad. 
 
 Considering that as an average for seven years, and for the last 
 year, from seven to ten of the large cities of the world fall within 
 the limit of the first class, it is painfully evident that a typhoid 
 fever death rate of 10 per 100,000 of population living is easy of 
 attainment when municipal corporations really desire to bring it 
 about. Few people realize the relative standing of cities in the 
 hygiene of their public water supplies ; and this classification will 
 enable a comparison to be made, which should convince even the 
 skeptical that in the matter of our public water supplies we are far 
 behind the larger cities of Europe. 
 
 Referring to the cities of the first and second class, Rotterdam 
 draws its water supply from the River Maas, one of the mouths of 
 the Rhine, and passes it through sand filters before it is delivered 
 to the consumers. Amsterdam* and The Hague draw their water 
 supplies from the sand dunes, and afterwards subject it to careful 
 filtration. Vienna and Munich depend upon water from large 
 springs at high elevation in the Alps ; the former in the Schnee- 
 berg, 65 miles southwest of the capital, and the latter in the Mang- 
 fall valley, 37 miles from the city. Dresden is supplied from an 
 infiltration well on the banks of the Elbe, which, according to Mr. 
 B. Salbach,f intercepts an underground flow parallel to the river. 
 
 Berlin, the largest city in the first class, draws its water from 
 
 * This city takes a portion of its water supply from the River Vecht. 
 t Transactions American Society of Civil Engineers, vol. xxx., p. 293. 
 
76 THE PURIFICATION OF WATER. 
 
 two sources, the Stralau Works from Lake Tegel,* an expansion 
 of the River Havel, and the Frederickshagen Works from Lake 
 Miiggel, an arm of the River Spree. At both stations the water 
 is subjected to sand filtration before it is pumped to the distribut- 
 ing reservoirs or mains. Some of the most conscientious and com- 
 plete investigations of sand filtration have been conducted with the 
 niters of these works by Plagge, Proskauer, Piefke, and others. 
 
 Either spring water or filtered well and river waters constitute 
 the sources for the cities in the first class. 
 
 London falls in the second class, and here the sources of supply 
 are the Rivers Thames and Lea, and wells in the chalk or soft 
 limestone. The enormous consumption of water by this city, now 
 quite 200,000,000 imperial gallons per day, compels an abstrac- 
 tion at times of quite 30 per cent of the whole stream flow of the 
 Thames, and even a larger proportion of the Lea. 
 
 The watersheds of these streams are heavily populated by urban 
 and rural communities ; and in spite of precautions to prevent sew- 
 age contamination of the water, there is no doubt that, with a low 
 stream flow and small dilution of sewage effluents, the pollution of 
 the raw river waters at times is very great. A review of the his- 
 tory of the London sand filters and their operation further suggests 
 that the same solicitude for water quality is not found here as in 
 some of the Dutch and German cities, and to these facts may be 
 ascribed the higher death rates from typhoid fever in London than 
 in the cities grouped in the first class. 
 
 The natural conditions of the water supplies of Edinburgh, New 
 York, and Brooklyn are in some respects alike, with the probability 
 that the New York watershed is more completely protected from 
 manifest sewage pollution than the watershed of Edinburgh in the 
 Pentland Hills. Filtration of the water is occasionally resorted to 
 by the latter city, according to Mr. J. P. Kirkwood, mainly to cor- 
 rect the turbidity ; and it is not known that any large improvement 
 has taken place in the manner of operating the filters of late years. 
 The hygiene of the water supply of Edinburgh should be like that 
 of New York, and the typhoid fever rates support this view. 
 
 * The works at Stralau have not been operated for several years, but are kept in reserve. 
 All water is now supplied from the Lake Miiggel Works. 
 
CLASSIFICA TION OF CITIES B Y TYPHOID FE VER STA TISTICS. 77 
 
 Hamburg, a city which for a period of seven years falls in the 
 second class, but which since the introduction of filtered water is 
 found in the first class, has had an unique experience. Prior to 
 May, 1893 (the spring following the cholera epidemic), the water 
 was drawn from the Elbe, and, apart from a limited sedimentation 
 in reservoirs of small capacity, was sent to the consumers with no 
 improvement in its quality. Since the date mentioned all water 
 has been filtered under the supervision of Dr. Dunbar, director of 
 the Hygienic Institute, and Mr. Rud Schroder, inspector of the 
 Water- Works, with the result that the typhoid rates have been re- 
 duced quite 73.5 per cent. No other change has been made in 
 the water supply than its filtration. The water of the Elbe, with 
 its sewage pollution from all sources above the city, is used now 
 as heretofore ; but none of it goes into the distributing mains until 
 it is first passed through an elaborate system of plain sand filters. 
 
 Cities which depend upon water supplies from sources known 
 to be sewage polluted, and of which no attempt is made at puri- 
 fication, very naturally suffer from high typhoid rates ; and an 
 examination of the cities in classes four to seven shows that nearly 
 all of them are in our own country. 
 
 The classification of cities for the last year of report (1896) 
 enables Hamburg, Newark, Jersey City, and Lawrence to show 
 what has been accomplished toward reduction of the typhoid rates 
 by the substitution of good for bad water. 
 
 With reference to the use of public water supplies for dietetic 
 purposes, in cities of Europe, the author, in a recent lecture upon 
 the hygiene of water, published in the Dietetic and Hygienic 
 Gazette, Philadelphia, October, 1896, says: 
 
 "When comparisons are made of the typhoid death rates of cities in 
 Europe with cities in this country, the claim is sometimes urged that the 
 people of Europe, and especially of Germany, are not water drinkers, that 
 beer is their usual beverage ; and upon the other hand, that the people of the 
 United States are not beer drinkers, but water drinkers, and therefore more 
 exposed to water-carried infections. When it is stated that *The Hague has 
 a typhoid death rate of 5 per 100,000 of population, this, according to some 
 critics, is not to be taken as an evidence of the high quality of water supplied 
 to the city, but as an indication that the people of The Hague generally do 
 not drink water, and depend upon beer or some other manufactured beverage." 
 
78 THE PURIFICATION OF WATER 
 
 " This expression of doubt by some, that water may be so purified by artifi- 
 cial means, or may be naturally so pure as to largely diminish the probability 
 of one contracting typhoid fever by drinking it, suggests inquiry along three 
 lines : 
 
 "1. Is the water supplied to certain foreign cities such as to reduce the 
 typhoid fever rates or inhibit the disease, if it were generally used as a bever- 
 age ? 
 
 " 2. Is the water generally used for drinking in the larger cities of the 
 United States such as to be the probable cause of our high typhoid fever 
 rates ? 
 
 "3. Is it true that the people of London, Berlin, Hamburg, and other 
 European cities are largely beer drinkers, while the people of Boston, New 
 York, Cincinnati, and other American cities are largely water drinkers? 
 
 " It is not possible to answer the first question directly. Despite the great 
 chemical and bacterial improvement by sedimentation and (or) filtration of 
 certain polluted waters like that of the Elbe at Hamburg, and the Maas at 
 Rotterdam, one cannot say positively that such waters, even after treatment, 
 will not contain the typhoid bacillus, or be the carrier of infection to some ; 
 and we are compelled to measure the improvement in quality of such waters 
 by their influence upon the health of the people who use them. 
 
 "In regard to the quality of water supplied to the people of certain cities 
 in Europe, it should be manifest, if this was not to be used as a drinking- 
 water, that a very large annual expense could be avoided in those cities by 
 pumping water direct into the reservoirs or street mains from any convenient 
 source, without attempting in any manner an improvement in its quality be- 
 fore it is distributed to the consumers. 
 
 " Water of high hygienic quality is not required for the sprinkling of 
 streets and lawns, for the extinguishment of fires, for the flushing of sanitary 
 apparatus, for steam boilers, and many other uses ; and if the water is not to 
 be used for drinking and other dietetic purposes, great care and expense in 
 the selection of a source of supply, or in efforts at improvement of the quality 
 of water, are surely wasted. 
 
 " Considering that over ninety-eight per cent of the consumption of water 
 by any large city is for purposes wholly unaffected by its hygienic quality, it 
 would seem very singular indeed that a city like Berlin (for instance) should 
 be at an extra expense of ten dollars per million gallons to fit the water for 
 drinking purposes, before a gallon of it is permitted to go into the public mains. 
 This great cost for purification of the water from Lakes Miiggel and Tegel is 
 not necessary if the water is to be used only for street sprinkling or other pur- 
 poses apart from drinking and the requirements of the cuisine. Moreover, the 
 water from Lake Miiggel, after it has passed through the filters at Fredericks- 
 hagen, as we have shown, is chemically and bacterially as pure as many nat- 
 ural spring or deep well waters which are known. to be altogether safe for 
 drinking purposes, and chemically and bacterially pure or nearly pure waters 
 
CLASSIFICA TION OF CITIES B Y TYPHOID FE VETTSttTIS TICS. 79 
 
 are not needed for any of the many uses of water, excepting for drinking, cook- 
 ing, and the washing of uncooked articles of diet. No one would propose an 
 elaborate and expensive treatment of a polluted water unless some portion of 
 it was to be drunk. 
 
 " The water of the River Elbe, when it reaches Hamburg, is of sufficient 
 purity for all ordinary purposes ; but the most modern works upon a large scale 
 for the improvement of polluted waters have recently been devised by that 
 city, and these works are carefully operated to reduce the noxious properties 
 of the Elbe water before it is distributed to the citizens. This work of purifi- 
 cation is not intended to make the water better for the great majority of the 
 uses to which it is applied, but to make it a water which the inhabitants can 
 drink with the least risk of infection from typhoid fever and other water-carried 
 diseases. If it were a fact, as some are disposed to think, that the people of 
 Hamburg do not drink water, why should that city be at such great effort and 
 expense to render the water of the Elbe fit to drink before it is permitted to 
 go to the consumers ? 
 
 " But the most pronounced efforts to procure a supply of public water which 
 certainly shall not be the cause of infection is found at The Hague, where the 
 water is first obtained from wells driven in the sand dunes and afterwards 
 passed through filters of sand, the grade of which is finer than that of nearly 
 every other city which has adopted sand filtration. The water of The Hague 
 as it comes from the driven wells in the dunes very likely is equal to that of 
 any of the driven well waters which we are accustomed to drink with a feeling 
 of perfect security. But the officials of that city, not content with a water 
 which at its source is far superior to nearly all of our public waters, set about 
 to improve its hygienic quality by slow filtration through beds of fine sand, with 
 the result that their city has had for many years nearly the lowest recorded 
 typhoid fever death rate of any of the large cities of the world. Are we to 
 ascribe this low typhoid rate to the drinking of beer, gin, or Schiedam schnapps 
 by the people of The Hague, or shall we credit it to the drinking of this excep- 
 tionally pure water from the public mains ? 
 
 " From such information as the author has been able to obtain, it is alto- 
 gether probable that in the consumption of beverages other than water we are 
 quite abreast of the people of this old Dutch city, and the only certain differ- 
 ence in the conditions surrounding the two cities which would affect the typhoid 
 rates is found in the quality of their respective public water supplies. 
 
 " It is not difficult to answer the second question. Nearly all the water 
 supplies of the large cities of this country are polluted with household sewage, 
 and are the carriers of the typhoid bacillus from the sick to the well. Having 
 knowledge of the fact that many of our large cities are daily drawing water for 
 drinking and other purposes from sources of known sewage pollution, it is 
 proper to look upon the typhoid rates of such cities as the natural result of this 
 indifference to one of the first laws of health, viz., a pure drinking-water. 
 
 " In regard to the third line of inquiry the author is not able to state the 
 
80 THE PURIFICATION OF WATER, 
 
 per capita per annum consumption of beer by many of the larger cities of 
 Europe; but the greatest consumption is accredited to Munich, which for one 
 year used 125 gallons per capita.* An investigation of the probable consump- 
 tion of beer by the larger cities of this country reveals the startling fact that 
 even the city of Boston consumed 65 gallons of beer per capita during the year 
 1894, while Cincinnati indulged itself to the extent of 80 gallons per capita, and 
 Milwaukee, for the same year, reached the respectable figure of 105 gallons 
 per capita. Of the list of ten of the larger cities of the United States, the low- 
 est per capita per annum consumption for the year 1894 was 46 gallons, and the 
 highest 105 gallons. The amount of beer made and drunk in the United States 
 for 1894 allows nearly 16 gallons for every man, woman, and child of the whole 
 population. We all know that large quantities of beer (and other artificial 
 beverages) are made and sold in this country, and we know that these are not 
 substituted for the industrial and sanitary uses of water. From the limited 
 information at command, I am sure it would be a mistake to assume that the 
 people of Europe drink nothing but beer,f or that the people of this country 
 drink nothing but water." 
 
 * Encyclop&dia Britannica, ninth edition, vol. xvii., p. 32. 
 
 f The average daily consumption of water by Munich, 1895-1896, was 12,947,683 U. S. 
 gallons, corresponding to a daily per capita consumption of 32.38 gallons. 
 
PURE AND PURIFIED WATERS. 81 
 
 f ' 
 
 CHAPTER VI. 
 
 PURE AND PURIFIED WATERS. 
 
 WATER supplies from sources of known purity undoubtedly are 
 superior to purified water from polluted sources ; but these are very 
 rare, and only a* few cities peculiarly favored, like Vienna, Munich, 
 and Dresden abroad, and some of the smaller cities and villages 
 in this country, can make them available. The water from certain 
 mountain springs and streams, and from some deep wells, from 
 the standpoint of hygiene, may be considered " pure ; " while that 
 supplied to cities where filtration or sedimentation in large reser- 
 voirs is practiced, may be regarded as " purified " water. 
 
 Viewed from a chemical and bacterial standpoint, there is no 
 degree to pure water ; but from the hygienic point of view there 
 may be, and apparently are, degrees of purity. The water of 
 Vienna is said to be naturally pure, so is the water of Manchester 
 (England) and New York ; but accepting the typhoid fever rates 
 as an index of water quality, the water of Vienna is by far superior 
 to that of either of the other two cities. Manchester and New 
 York attempt to protect the drainage grounds of their sources, 
 and preserve the water from direct sewage pollution, which efforts 
 are only partially successful. Vienna, Munich, and a few other 
 cities seek their water supplies in sources which apparently are 
 beyond the reach of pollution. 
 
 Liverpool, like New York and Boston (new works), has sought 
 its water supply in a district which is sparsely inhabited, and not 
 exposed to the sewage from large organized communities ; and as 
 an additional precaution in behalf of the public health, provision 
 is made for filtration of this water before it reaches the city. 
 
 In considering sources of water supply in mountain springs at 
 moderate distance from cities to be supplied, and sources in deep 
 wells, it should not be overlooked that similar sources cannot be 
 
82 THE PURIFICATION OF WATER. 
 
 made available for all or even many cities. Nor should the fact 
 be ignored, that the enormous per capita consumption in nearly 
 every American city renders the problem of a " pure " or "puri- 
 fied" water supply for our cities much more difficult of solution 
 than in the cities of Europe. Berlin, with about the same popula- 
 tion as Chicago, uses less- than one-fourth of the quantity of water 
 per diem ; London, with a population of over 5,000,000, probably 
 uses no more water than Philadelphia ; while Hamburg, with nearly 
 twice the population of Cincinnati, uses less than three-fourths as 
 much water. The consumption or rather the waste of water in 
 many cities, is a serious impediment to improvement in works of 
 public water supply ; and the abuse of water privileges must be 
 curbed, if we are to have water of the same quality as that of many 
 of the cities of Europe. 
 
 Of the larger cities of the United States which derive their 
 water supplies from driven wells, Brooklyn thus obtains from 
 many sources, covering a large territory, about 32,000,000 gallons 
 per day,* or over four-tenths of the daily supply. The maximum 
 yield of the system of artesian wells at Memphis, Tenn., has been 
 stated at 16,000,000 gallons per day.f The maximum capacity of 
 the system of driven wells at Dayton, Ohio, is given as 6,750,000 
 gallons, and at Lowell, Mass., as 12,000,000 gallons per day.J 
 Upon the same authority the average daily consumption of water 
 from the system of driven wells at South Bend, Ind., is 1,900,000 
 gallons ; and the maximum daily consumption from the artesian 
 wells at Jacksonville, Fla., is given as 1,557,557 gallons. Many 
 smaller cities have systems of artesian or non-flowing wells which 
 yield from a few hundred thousand to one or two million gallons 
 per day, and all such may be regarded as highly favored by nature 
 in the matter of their public water supplies. 
 
 In Europe, especially in Germany, it is the policy to seek 
 public water supplies in sources of natural purity, such as moun- 
 tain springs and deep wells, where these are available, and to 
 
 * Mr. I. M. DeVarona, in Report on the Future Extension of the Water Supply of the 
 City of Brooklyn, 189(5, p. 26. 
 
 f Report on Extension and Betterment of Cincinnati Water-Works, 1896, p. 27. 
 J Manual of American Water-Works, 1897. 
 
PURE AND PURIFIED WATERS. 83 
 
 limit the consumption of water to the yield 'of such sources. 
 When the yield, as at Dortmund for instance is relatively large, 
 the allowance per capita per diem is correspondingly liberal ; while 
 at Leipsic, where the yield of ground water is relatively small, 
 the per capita consumption also is small. 
 
 In order to utilize to the fullest extent the natural sources of 
 pure water for public supply, the people of Germany are willing 
 to limit the use and waste of such water sometimes to very small 
 per capita daily allowances, reasoning, doubtless, that the require- 
 ments of hygiene are better satisfied with small amounts of pure 
 water than large volumes of polluted water. 
 
 Among the larger cities of Europe which depend partly or 
 wholly upon ground waters may be mentioned London, the Kent 
 Works of which during July, 1896, supplied from deep wells a 
 daily average of 23,270,000 U. S. gallons to an estimated popula- 
 tion of 583,436, allowing thus nearly 40 gallons per capita. 
 
 In the table on the following page are given the principal 
 cities of Germany, etc., which depend in whole or part on ground 
 water supplies.* 
 
 If the double system of water supply which prevails in parts of 
 Paris (where the very excellent water of the Vanne is used for 
 dietetic purposes) should be adopted by cities in this country, 
 then it will in most instances become a comparatively easy task to 
 secure the limited quantity of water required for drinking and culi- 
 nary uses, either from sources of known purity or by very careful 
 filtration. 
 
 The great advantage of water from a source not open to sew- 
 age or semi-sewage pollution, as, for instance, deep wells intercept- 
 ing water which has been thoroughly purified in passing through 
 the drift, over water purified by any process of filtration, is found 
 in the fact that such water is at all times safe ; whilesafety to the 
 public of purified water depends altogether upon the skill and care 
 of the officials in charge of the filters. A lack of technical knowl- 
 edge or vigilance upon their part may result in great damage to 
 the health of the people supplied. 
 
 Naturally "pure" water is not available by the great majority 
 
 * Statistische Zusammenstellung der Betriebs Ergebnisse von Wasser-werken, Munich, 1895. 
 
84 
 
 THE PURIFICATION OF WATER. 
 
 of cities. To supply a so-called pure water to London from 
 sources in Wales, nearly $200,000,000 will be required ; and even 
 in that instance it is proposed by the County Council to filter the 
 
 GERMAN CITIES SUPPLIED WITH GROUND WATER. 
 
 CITIES. 
 
 DATE 
 
 OF 
 
 REPORT. 
 
 POPULATION. 
 
 DAILY YIELD 
 OF WELLS. 
 U. S. GALS. 
 
 PER CAPITA 
 DAILY 
 ALLOWANCE. 
 U. S. GALS. 
 
 PROPORTION 
 OF WATER 
 FROM WELLS. 
 
 Schalke, 
 
 1893 
 
 280,000 
 
 11,429,790 
 
 40.8 
 
 All. 
 
 Dortmund, 
 
 1893-4 
 
 170,000 
 
 9,617,260 
 
 56.6 
 
 " 
 
 Cologne, 
 
 1893-4 
 
 286,000 
 
 8,425,535 
 
 29.5 
 
 " 
 
 Dresden, 
 
 1893 
 
 309,000 
 
 6,912,000 
 
 22.4 
 
 
 
 Bochum, 
 
 1893-4 
 
 148,500 
 
 6,764,100 
 
 45.5 
 
 it 
 
 Leipsic, 
 
 1893 
 
 391,000 
 
 6,131,170 
 
 15.7 
 
 u 
 
 Stockholm, 
 
 1893 
 
 251,000 
 
 6,045,850 
 
 24. 
 
 U 
 
 Copenhagen, 
 
 1893 
 
 337,500 
 
 5,990,789 
 
 17.8* 
 
 
 Essen, 
 
 1893-4 
 
 135,000 
 
 4,869,495 
 
 36. 
 
 All. 
 
 Barmen, 
 
 1893-4 
 
 125,000 
 
 4,380,533 
 
 35. 
 
 U 
 
 Augsburg, 
 
 1893 
 
 80,000 
 
 4,253,900 
 
 53.2 
 
 " 
 
 Dusseldorf, 
 
 1893-4 
 
 157,000 
 
 4,204,468 
 
 27. 
 
 U 
 
 Charlottenburg, 
 
 1892-3 
 
 255,000 
 
 4,204,987 
 
 15.8 
 
 U 
 
 Elberfeld, 
 
 1893-4 
 
 144,000, 
 
 3,906,663 
 
 27.1 
 
 U 
 
 Miilheim, Ruhr, 
 
 1893-4 
 
 67,000 
 
 3,729,280 
 
 55.7 
 
 it 
 
 Hannover, 
 
 1893-4 
 
 221,000 
 
 3,477,132 
 
 15.7 
 
 II 
 
 Freiberg, 
 
 1893 
 
 51,000 
 
 3,157,980 
 
 62. 
 
 " 
 
 Frankfort, 
 
 1893 
 
 198,800 
 
 3,057,138 
 
 . . . 
 
 40$ 
 
 Duisburg, 
 
 1893-4 
 
 73,875 
 
 2,793,280 
 
 37.8 
 
 All. 
 
 Carlsruhe, 
 
 1893 
 
 80,000 
 
 2,7.63,514 
 
 34.5 
 
 ii 
 
 Halle, 
 
 1893-4 
 
 123,000 
 
 2,605,294 
 
 21.1 
 
 a 
 
 Crefeld, 
 
 1893-4 
 
 106,000 
 
 2,301,588 
 
 22.0 
 
 II 
 
 Witten, 
 
 1893-4 
 
 45,000 
 
 2,195,849 
 
 48.8 
 
 II 
 
 Strasburg, 
 
 1893-4 
 
 95,000 
 
 1,964,511 
 
 20.6 
 
 ti 
 
 Vienna, 
 
 1893 
 
 998,000 
 
 1,877,795 
 
 
 10# 
 
 Upon this aspect 
 
 water before it is delivered to the consumers, 
 of the water question Dr. Leff man f says : 
 
 " When propositions for nitration are made, it is usual for some persons 
 to suggest that a pure water supply should be selected. . . . Surface water 
 is so liable to pollution that the word pure ' has, in regard to it, only a com- 
 parative sense ; and in establishing an elaborate water supply, we should estab- 
 lish systems of storage and nitration, no matter how excellent may be the 
 district in which the water is collected and through which it flows." 
 
 * Springs and wells, proportion not given. f Public Health, 1897, p. 118. 
 
PURE AND PURIFIED WATERS. 85 
 
 Changes may occur in the quality of water from natural sources 
 by subsequent pollution of the tributary watersheds. Large water- 
 sheds constituting the sources of supply for cities are more exposed 
 than small watersheds used by villages, and absolute security in 
 either case is to be had only by complete control of the effective 
 drainage grounds. 
 
 The water of deep wells may become polluted by sewage from 
 improvements which encroach on their drainage area ; and the area 
 drained by such wells should be free from habitation, and such 
 commercial operations as are calculated to contaminate the soil and 
 pollute the rainfall which percolates through the soil to the wells. 
 
 A proposition by the author (1894) to sterilize by distillation, 
 and distribute through a separate system of mains, that portion of 
 a city water supply which was used for drinking and dietetic pur- 
 poses, developed much adverse criticism. Partly, as alleged, be- 
 cause such water, when deprived of the minerals in solution, would 
 not be so favorable to the animal system as are natural waters ; 
 partly because of the difficulty of educating people to the use of 
 sterilized water when other water was less expensive and more 
 convenient to obtain; and partly because of the expense of a dupli- 
 cate system of mains to distribute such water to the consumers. 
 
 The first objection is the only one worthy of serious considera- 
 tion. If it is true that sterilized water exerts a prejudicial influence 
 on digestion or any of the animal functions, then it should not be 
 recommended ; but observation among people who are regularly 
 using distilled water does not bear out the assertion sometimes made, 
 that such water is less beneficial for dietetic uses than clean cistern 
 or pure well water. Neither is the author aware that systematic 
 experiments have ever been conducted to ascertain the real influence 
 of sterilized water on the human system, unless it may be held 
 that the favorable results of the use of distilled water in the United 
 States navy and on ocean steamships furnishes the desired data. 
 If the salts and minerals lost by distillation are really essential to 
 a perfect drinking-water, it would seem to be much safer to add 
 these in proper proportions to distilled water, than to assume the 
 risk to health which accompanies the indiscriminate use of natural 
 waters for drinking. 
 
86 THE PURIFICATION OF WATER. 
 
 If it were true that a distilled water, lacking in lime, soda, 
 potash, etc., was unfitted for the manufacture of teeth and bone in 
 young children, this fact should be manifest in the children of 
 suburban and other villages, which depend almost entirely for 
 drinking upon cistern water. The author's observation for many 
 years along this line furnishes no proof that cistern water, if clean 
 and free from objectionable organic matter and bacteria, is not a 
 perfectly safe and satisfactory drinking-water. In fact, rain-water 
 falling in suburban districts, caught on clean slate roofs, and col- 
 lected in clean cisterns, should furnish the purest and best of 
 natural waters. Underground cisterns intended for the collection 
 and storage of drinking-waters should be tight, and not subject to 
 contamination through the soil from neighboring cesspools and 
 vaults ; otherwise, pollution may occur, and such cistern water 
 would be quite as objectionable in a hygienic view as any other 
 sewage-polluted water. 
 
 Certain spring and well waters may be quite free from organic 
 matter and bacteria, and still be dangerous to health by reason of 
 minerals in solution. Lead, arsenic, copper, etc., and iron in excess, 
 are objectionable ingredients of drinking-waters ; and petroleum 
 above one part in two millions, it has been stated, unfits water for 
 drinking.* Some of these substances may be in waters which, when 
 tested by bacteriological methods and the microscope, would be 
 found very pure. Chemistry, however, can reveal and measure them. 
 
 While certain surface waters can be carried in large, deep res- 
 ervoirs with an improvement in quality, water from ground sources 
 cannot be stored in open shallow reservoirs without developing a 
 growth of vegetable and animal matter, the luxuriance of which 
 depends upon the climate and sunlight, and to some extent upon 
 the mineral constituents of the water. Tall tanks and stand-pipes, 
 the depth of which is usually greater than that of earthen reser- 
 voirs, and the water area small in relation to capacity, may be 
 used to store ground waters without apparent change in quality ; 
 but in these the quantity of water stored is always small when 
 compared with the daily consumption. 
 
 * Water Supply, Chemical and Sanitary, Wm. Ripley Nichols, New York, 1883, p. 75. 
 
PURE AND PURIFIED WATERS. 87 
 
 The investigations of Mr. G. C. Whipple, of*the Massachusetts 
 State Board of Health,* show very conclusively that sunlight is 
 the controlling factor in the development of algous growths in 
 stored waters, which suggests that the light should always be 
 rigorously excluded from reservoirs and large tanks intended for 
 the reception of ground waters. It has been the author's usual 
 practice to cover steel tanks and towers for the storage of ground 
 waters, but he is not aware of any trouble or complaint arising 
 from the storage of ground waters in the few open tanks of works 
 with which he has been associated. 
 
 According to Professor Mason, f " to keep a ground water in 
 good condition it is necessary to cover the reservoir. Such waters 
 are usually charged with mineral matter suitable for plant food, 
 and the higher organisms will be likely to grow therein unless 
 light be excluded." 
 
 The rapid development, during the summer, of vegetable 
 growths in shallow open reservoirs carrying ground waters, is well 
 known ; and the growth and (or) decay of some of these organisms 
 have produced unpleasant tastes and odors in stored waters, but no 
 proof is at hand that they have been the specific cause of disease. 
 Numerous investigations have shown that asterionella, nostoc, oscil- 
 laria, and other of the green algce, crenotJirix (fungi) and uroglena 
 (infusoria), have each at times imparted peculiar tastes or odors to 
 stored waters ; and it is altogether likely that other of the micro- 
 scopic organisms in water may be concerned in producing effects 
 which justify objections to the use of such waters from an aesthetic 
 standpoint, even if not positively injurious to health. 
 
 Assuming that the development of vegetable organisms in 
 stored waters depends principally upon the penetration of light, 
 it is obvious why turbid river and surface waters can be stored in 
 large, deep reservoirs for great lengths of time without injury, and, 
 as a rule, with positive improvement in their quality. During the 
 early days of storage the color is so strong, and the water so nearly 
 opaque, that there is no penetration of light ; and upon subsidence 
 of the heavier matters in suspension many of the vegetable organ- 
 
 * Journal of New England Water Works Association, September, 1896. 
 t Water Supply, by William P. Mason, New York, 1896, p. 261. m 
 
88 THE PURIFICATION OF WATER. 
 
 isms, including the bacteria, are carried down, and deposited on the 
 bottom and slopes of the reservoirs. Upon the other hand, the 
 usual properties of ground water, viz., 
 
 (1) Limpidity and lack of color, 
 
 (2) Small or no organic matters in suspension, 
 
 (3) Large amounts of dissolved salts readily assimilable by 
 plant life, are favorable to the growth of cryptogams. 
 
 Add to these light and heat by exposure in open reservoirs, 
 and all the conditions are present essential to the rapid growth of 
 algae. The Vanne water, which is used for dietetic purposes in 
 portions of Paris, is obtained from springs ; and to preserve it 
 without change of quality it is conducted to the city in a closed 
 conduit, received in distributing reservoirs from which the light 
 is carefully excluded, and reaches the consumer quite as pure as 
 it was upon issuing from its mountain source. 
 
 The smaller cities and villages often are peculiarly favored in 
 sources of satisfactory public water supply ; while larger cities, 
 where the consumption of water per capita and in the aggregate 
 is greatly in excess of that of the smaller communities, are 
 compelled to procure the required daily volumes of water, in 
 many instances, from sources utterly unfit for domestic uses. A 
 mountain spring or system of driven wells, which will furnish an 
 abundance of pure water to some small municipality, would be 
 too insignificant for consideration as a source of water supply 
 to a city ; and intelligent people are prone to neglect the fact that 
 a water source which may meet the requirements of a village of 
 a few thousand population would be inadequate to supply even 
 the drinking and culinary water of a large city. 
 
 Referring to present sources of city water supply, it can be 
 said that Vienna, Munich, Dresden, Stockholm, Copenhagen, a 
 portion of Paris and other cities abroad, together with several of the 
 smaller cities and many villages in this country, have water sup- 
 plies which are naturally pure ; while London (omitting the Kent 
 Works), Berlin, Hamburg, Liverpool, and many smaller cities in 
 Europe, Lawrence, Mass., and several other cities in this country, 
 have purified water supplies. 
 
PUKE AND PURIFIED WATERS. 89 
 
 
 
 In this connection it is difficult to discuss the water supplies of 
 cities in America using mechanical filters. The typhoid fever rates 
 of these cities are considerably higher than the rates of cities in 
 Europe which use plain sand filtration for the purification of 
 polluted waters ; and inquiry among the manufacturers of the 
 filters, and the officials of the water-works using them, reveals the 
 startling information that (in instances at least) the filtered water 
 is not generally used by the people of such cities. This is also 
 true of certain cities and villages which have very excellent water 
 supplies from natural sources. 
 
 In a certain city of Ohio, where the public water supply is from 
 a system of driven wells and of very excellent quality, it is said 
 that less than one-fourth of the population draw their water supply 
 from the city mains. The remainder, after nearly thirty years' 
 experience with a public system of water- works, still depend for 
 domestic purposes upon the water from wells and cisterns. In the 
 interest of the public health, and where the public water supply is 
 unexceptional, the use of wells and cisterns for drinking-water 
 should be prohibited (as they are in Vienna), and the use of the 
 public water be made compulsory. Of what avail is it to secure 
 water of high quality if the people are permitted afterwards to 
 take their drinking and dietetic water from sources of doubtful 
 value, and, as is well known to many sanitary officers, from sources 
 exposed to sewage pollution ? 
 
 It is a strange anomaly that in the larger cities (many of which 
 are supplied with sewage-polluted waters) all the people are by 
 the force of circumstances compelled to take their water from the 
 public mains, while in the small cities and villages, in which the 
 public water is often of most excellent quality, the use of the better 
 water is altogether optional with the people. 
 
 It is impossible to tell how much physical suffering might be 
 traced to water supplies, which are regarded by municipal corpora- 
 tions and water companies as fit for domestic uses, and to the 
 continuous use by cities of waters of known sewage pollution. 
 Certain it is that typhoid fever, cholera (in the Orient), and other 
 intestinal disorders, annually claim thousands of victims, which 
 would be saved if all people were equally and sincerely interested 
 
90 THE PURIFICATION OF WATER, 
 
 in having drinking and dietetic waters of the highest attainable 
 purity. 
 
 In the mountains, where the population is sparse, water of sat- 
 isfactory quality and in abundance is often found. But at lower 
 elevations, along the rivers and lakes, and at tide water, where the 
 people are collected in large numbers, mountain sources of water 
 supply are rarely available. Doubtless " there is enough whole- 
 some water on the face of the earth to supply all the inhabitants 
 thereof," but the conditions clearly demonstrate that this is not 
 well distributed. 
 
 Cities have been located, not according to the rules of hygiene, 
 but according to the requirements of commerce. Revenue has 
 been the dominating factor ; and upon the sanitarian has fallen 
 the burden of rectifying evils which have followed the total disre- 
 gard in so many cases of the fundamental laws of health. 
 
 In the introduction to this work a statement is made that 
 "water is an essential of human existence;" and this is true, not 
 only as it is used in connection with the body needs, but in con- 
 nection with the fruits of the soil, in tempering the atmosphere 
 and heat of the earth's crust, and in many other ways. Restricted 
 to our animal requirements, it may be said that " pure water " is 
 "an essential of health," while "impure water" involves hazards 
 to life and health which we have no right to incur. 
 
 It is feasible for a small percentage of the world's population 
 to procure water supplies from natural sources which will satisfy 
 the most advanced requirements of hygiene, but for the great 
 majority of the people satisfactory water supplies are obtainable 
 only by works of artificial purification. 
 
CITATIONS ON TYPHOID FEVER EPIDEMICS. 91 
 
 CHAPTER VII. 
 
 CITATIONS ON TYPHOID FEVER EPIDEMICS. 
 
 PRACTICAL illustrations, so far as they are available, concerning 
 the causes of typhoid fever epidemics, are especially valuable in 
 supporting the theory that a sewage-polluted water, or a water 
 carrying the germs of typhoid fever, will produce infection of the 
 same disease in persons who may drink such water, or who in 
 some way may have ingested food which has been in contact with 
 such water. In this connection it is not necessary that a line 
 should be drawn between the colon bacillus and the typhoid bacil- 
 lus as found in the human spleen ; either or both may be going 
 into the sewers of cities, and from the sewers into rivers, lakes, 
 and other sources of public water supply. It is sufficient and 
 prudent to assume that the dejections of a typhoid patient may 
 contain the specific organism which is the cause of this disease, 
 and in the instances noted in this chapter the evidence is at times 
 overwhelmingly in support of this view. 
 
 LAUSEN, SWITZERLAND. 
 
 The typhoid epidemic which occurred in this village during the 
 latter part of 1872 has been mentioned so frequently and with so 
 much respect as to make it a classic in the epidemiology of this 
 disease. The circumstances were briefly as follows : A few cases 
 of typhoid fever, occurring at a distance of one or two miles 
 from the village, were supposed to have caused a contamination of 
 the water of the village well, from which an infection of 130 per- 
 sons, with 8 fatalities, in due time followed. From the official 
 report of the epidemic by the health officer of Basle, it appears 
 that a brook, passing near the premises where the original cases 
 of typhoid occurred, was used for the washing of the linen of the 
 patients, and at the same time as a channel for the disposal of 
 
92 THE PURIFICATION OF WATER. 
 
 the typhoid dejections. The course of this brook caused it to 
 join another stream which passed through Lausen at a considera- 
 ble distance below the village ; but for some time it had been sus- 
 pected that an underground connection existed between the brook 
 at a place below the original location of the fever and the village 
 well, although from surface indications, any sewage discharged 
 into the brook would pass into the larger stream below the village, 
 and not be the cause of pollution to the village water or create an 
 offense to its population. 
 
 After the epidemic, an investigation showed that the brook 
 did have an underground connection with the village well ; and 
 notwithstanding the percolation of the water through a mile or 
 more of pervious material, typhoid germs, which came into the 
 brook from the source of the original infection, were carried into 
 the public well, and spread the disease in the village at such an 
 alarming rate that one in every six of the whole population was 
 attacked. The original case of typhoid fever was held to have 
 been imported into the neighborhood of Lausen. 
 
 A resumJ of the simple facts in this instance shows : 
 
 1. No typhoid fever had been known of in Lausen for sixty 
 
 years, and the general health of the people was excep- 
 tionally good. 
 
 2. During the summer of 1872, several cases of typhoid had 
 
 occurred outside of the village, at a distance of a mile 
 or more. 
 
 3. The Furlen Brook was used as a channel of discharge for 
 
 the dejections of the original patients, and for the 
 washing of their linen. 
 
 4. Salt, thrown into the brook below the original location of 
 
 the disease, in due time appeared in the water of the 
 Lausen public well. 
 
 5. The infection was limited to those who drank of the water 
 
 of the public well, while families which abstained from 
 the use of the water of this well were not affected. 
 
 The Lausen circumstance has been so well described in certain 
 books and reports upon the subject, that it is here introduced 
 
CITATIONS ON TYPHOID FEVER EPIDEMICS, 93 
 
 t 
 
 principally to show that natural filtration so-called, through the 
 pervious materials of the drift, while it may render water very fair 
 to look upon, cannot be accepted as a safeguard against typhoid 
 infection, if the cause of infection should exist in the locality of a 
 source of ground water supply such as was had in Lausen. 
 
 CATERHAM, ENGLAND. 
 
 The outbreak of typhoid fever at Caterham occurred in the 
 early part of 1879, and was reported on early in April of that year. 
 In this instance there were upward of 100 attacks of the disease, 
 of which 19 were fatal. It was proved beyond doubt (according 
 to Mr. Edward Easton, Chairman of the Caterham Water Com- 
 pany), that the origin of the epidemic was due to the evacuations 
 of one workman accidentally contaminating the water of the well. 
 This workman, although suffering with a " walking " case of ty- 
 phoid fever, was employed in the tunnels connecting the wells 
 which constituted the source of water supply for Caterham and 
 Redhill. While so employed, his frequent dejections were col- 
 lected in a bucket, and from time to time hoisted out of the well. 
 During one trip of the bucket an accident occurred, causing a por- 
 tion of its contents to be spilled into the water ; and from this in 
 due time grew an epidemic of typhoid fever, alarming in propor- 
 tion to the population of the district supplied from this source. 
 
 Of this unfortunate occurrence, Dr. Thorne Thorne, who con- 
 ducted the investigation,* says : 
 
 " The water supply of Redhill and Caterham was derived from deep wells 
 in the chalk and lower greensand. It was discovered that a workman em- 
 ployed in excavating an adit between two of the wells had previously con- 
 tracted enteric (typhoid) fever at Croydon, and had been overtaken by diarrhoea 
 on several occasions while working in the well. There appeared no doubt 
 that the poison of the excreta was conveyed to the drinkers of the well water, 
 and communicated the disease." 
 
 The circumstances of the affair are revolting in the extreme ; 
 and how a person suffering with the early symptoms of typhoid 
 could be permitted to follow his occupation immediately in the 
 
 * " Report Royal Commission on Metropolitan Water Supply," Appendices to Minutes of 
 Evidence, p. 533. 
 
94 THE PURIFICATION OF WATER. 
 
 presence of the water from which a small population drew its daily 
 supply, is beyond comprehension. 
 
 Aside from a specific cause of disease which may be traced to 
 this lapse of moral responsibility upon the part of some one, it is 
 disgusting to know that a man would be so far lost to the sense of 
 decency as to continue his labors under conditions like these. Here 
 was a 'system of wells from which a large number of people were 
 daily drawing water for drinking and other purposes, in which 
 some repairs or improvements were being conducted while the 
 water was being pumped out of them for domestic consumption, 
 and a workman known to be suffering with diarrhea, if nothing 
 worse, was permitted to follow his labors in the wells, and dispose 
 of his dejections in a manner which was offensive to sentiment, 
 even if proper precautions had been observed to prevent contami- 
 nation of the water. But such precautions were not observed ; and 
 in due time the bucket was tipped, some of its contents went into 
 the water, and upwards of 100 persons were made to suffer as a 
 consequence of a proceeding which was little short of a crime. 
 
 In the exhaustive examination conducted by the Royal Com- 
 mission on Water Supply for London (1893), this Caterham oc- 
 currence is mentioned several times by the witnesses ; and the 
 principal deduction drawn from it by the Commission was, that if 
 the dejections of this workman had been tipped into the River 
 Thames instead of the Caterham wells, it would have been so 
 largely diluted as to have had no effect on the health of the people 
 who take their water from the Thames companies. 
 
 Stripping the matter of all verbiage, it appears that 100 cases 
 of typhoid fever with 19 deaths were traced to the contamina- 
 tion of this Caterham water, by the accidental discharge of a portion 
 of the dejecta from a mild case of typhoid fever. 
 
 PLYMOUTH, PA. 
 
 The town of Plymouth, Pa., in the spring of 1885 was visited 
 by a severe epidemic of typhoid fever, which had its origin in the 
 pollution of the public water supply by the dejections of a single 
 isolated patient. 
 
 The water of the town is impounded in a reservoir at an eleva- 
 
CITATIONS ON TYPHOID FEVER EPIDEMICS. 95 
 
 * 
 
 tion sufficient to maintain a supply by gravity ; and upon the drain- 
 age ground of this reservoir, at the time of and before the epi- 
 demic, lived a family, a member of which was temporarily located 
 in Philadelphia. This member came home ill with typhoid fever 
 during the winter of 1884-1885 ; and to avoid infection of the 
 family vault, his dejections were thrown upon the snow or ice-cov- 
 ered ground some distance from the residence. When the warm 
 rains of late winter set in, the snow and ice about the premises 
 melted, and ran into the reservoir. Within two or three weeks an 
 epidemic of typhoid fever occurred in Plymouth ; and out of a pop- 
 ulation of about 8,000, over 1,100 were stricken with the disease, 
 of which 114 were fatal cases. 
 
 In the Lausen epidemic, the typhoid germs went into the 
 water of the Furlen brook ; the water of the brook, or a portion of 
 it, went into the village well. Result, an epidemic of typhoid 
 fever in a little village which for more than two generations had 
 been wholly exempt from this disease. 
 
 In the Caterham and Redhill epidemic the dejections of a 
 typhoid patient were mixed with the water of a well which con- 
 stituted the source of supply for a small population ; this water 
 was pumped to the patrons of the water company, and an epi- 
 demic of typhoid fever followed. 
 
 In the Plymouth epidemic, the typhoid fever dejections were 
 thrown upon the frozen soil ; rains followed, and the runoff of 
 rainfall and melted ice carried these into the reservoir which con- 
 stituted the water supply for the city. Result, a typhoid epidemic 
 which at the time attracted the attention of the whole country. 
 
 In connection with the epidemic at Plymouth, Pa., the following 
 from the Engineering Record, July 18, 1896, will be of interest : 
 
 " The report was made public last week of Dr. Chas. P. Knapp, County 
 Inspector of the Pennsylvania State Board of Health, of Wyoming, Pa., who 
 in April last investigated the water supply of Plymouth, Pa., where typhoid 
 fever has been prevalent for some time past. Dr. Knapp reported that the 
 water of Plymouth is subject to contamination from a dairy farm on the head 
 waters of the stream from which the supply is taken. This is the same farm- 
 house and stream to which was traced in 1885 the typhoid epidemic in Plym- 
 outh, when there were 1,500 cases and 150 deaths."* 
 
 * The cases and deaths from this epidemic have been differently reported by different 
 authorities. 
 
96 THE PURIFICATION OF WATER. 
 
 ZURICH, SWITZERLAND. 
 
 " In the spring of 1884, Zurich was visited by a virulent outbreak of 
 typhoid fever, which, beginning in the month of March, reached its maximum 
 intensity in April, and practically disappeared by the end of June, fully two 
 per cent of the population having been attacked and nine per cent of the 
 cases proving fatal. The Commission appointed to inquire into the cause of 
 this epidemic arrived at the conclusion that the infection could not be traced 
 to abnormal meteorological or sanitary conditions ; but that the filtered Lim- 
 mat water, although clear and chemically satisfactory, contained an abnormal 
 quantity of bacteria. It was subsequently discovered that, owing to the dredg- 
 ing operations in connection with the new quay-works, the impure matter 
 deposited at the bottom of the river had been stirred up, and, not being effec- 
 tually retained by the submerged filter-bed, had found its way into the con- 
 crete main, which was by no means water-tight, and had been damaged during 
 the blasting and removal of an erratic block from the bed of the Limmat. 
 The reason why the defective condition of both filter-bed and conduit did not 
 arouse suspicion until after the outbreak of the epidemic was twofold, first, 
 the Limmat water was generally so clear and chemically pure that filtration 
 was considered of altogether secondary importance ; and, secondly, the per- 
 meability of the concrete, laid at 2 to 3 meters depth below the river bed, was 
 regarded as an advantage rather than a defect, inasmuch as the sand in which 
 the conduit was imbedded was supposed to act as a 'filtering medium. The 
 bacteriological investigations, in conjunction with the proved percolation of 
 impure matter through the filter and conduit, stamped the Limmat water as the 
 vehicle of infection. . . . Hence, at Zurich, the necessity was promptly rec- 
 ognized of providing an entirely new water supply for the town and suburbs."* 
 
 (The submerged filter mentioned above has been experimented 
 with by several cities in this country. It is a filter constructed in 
 the bed of the river, of such materials as may be convenient, with 
 no provisions for cleaning or renewal of the filtering materials. 
 Such filters, wherever used within the author's knowledge, have 
 been failures as sanitary devices.) 
 
 SPRING WATER, N.Y. 
 
 Spring Water in 1889 was a village of 600 population, situated 
 two and one-half miles south of Hemlock Lake, which constitutes 
 the source of water supply for Rochester, N.Y. This village was 
 on the watershed of the lake, and typhoid fever occurring there 
 
 * Preller on the Zurich Water Works, London, 1892, p. 7. 
 
CITATIONS ON TYPHOID FEVER EPIDEMICS. 97 
 
 was a menace to the health of the larger comrnunity. The water 
 supply of the village was obtained from open and driven wells 
 sunk a short distance into the drift. 
 
 A so-called endemic of typhoid fever, which occurred in this 
 village during October and November of that year, was made the 
 subject of careful investigation by Mr. George W. Rafter (then 
 assistant engineer of the Rochester Water- Works) and Dr. M. L. 
 Mallory. The following information is taken from the Report of 
 these gentlemen to the chief engineer of the water-works : 
 
 " The earliest clearly defined case of typhoid fever we found to be that of 
 ... a boy thirteen years of age, who, when taken sick with the disease on 
 Sept. 29, was employed at Snyder's Hotel. . . . Not only is the well at this 
 place in close proximity to the privy (thirty feet away), but half-way between 
 the well and the privy we found a board slop drain, which undoubtedly dis- 
 charges into the well a considerable portion of its contents. The family 
 claimed, however, that the water of this well had been considered bad for a 
 year and a half, and none of it had been used for domestic purposes during 
 that time ; the water so used having been all obtained from the well on the 
 adjoining premises. . . . We found the pump in working order, with pail 
 beneath the spout partly filled with water, and with a dipper in the pail. On 
 questioning the servant girl, it appeared very evident that the water was some- 
 times used. . . . 
 
 " Our view as to the origin of these cases of typhoid fever in the village 
 . . . is therefore as follows: The hotel was certainly an original center of 
 infection as ... four persons living there were taken sick with the disease ; 
 and while we are unable to establish the fact definitely, we consider it very 
 probable that some ' walking ' case of typhoid fever stopped at the hotel, and 
 . . . inoculated the hotel privy with germs of typhoid contained in the dejec- 
 tions. The chemical analyses of the water of the hotel well, . . . and the 
 bacteriological examinations, both show the water to be exceedingly bad, 
 unfitted for domestic use, and the environment such as to lead, with the cer- 
 tainty of a mathematical demonstration, to the conclusion that there is gross 
 pollution from the privy and slop drain." 
 
 The village of Spring Water at the time of this outbreak had a 
 population of 600, of which 20 were attacked with typhoid fever. 
 " The soil upon which the village stands is of an open, pervious 
 character for a depth of 10 to 20 feet," while the wells (open and 
 driven) had a depth of 10 to 18 feet. Previous to this attack 
 of typhoid fever the village had a record of being an unusually 
 healthful locality. 
 
98 THE PURIFICATION OF WATER. 
 
 It is stated in the report that the Eberth bacillus (b. typhosus) 
 was found, upon examination by Dr. H. C. Ernst of Harvard Uni- 
 versity, in a sample of water from one of the wells supposed to 
 have been concerned in spreading the infection. 
 
 The admixture of filth and drinking-water, which was assumed 
 to have been the secondary cause of this epidemic, is going on on 
 a very large scale all over the land, with this difference, however, 
 that in Spring Water the mixing was done right on the premises 
 of the citizens, while the plan usually pursued is to pollute the 
 water supply of one city with the sewage of another. Will a time 
 arrive when there will be a law on the Federal statute book which 
 will give the Engineer Corps of the War Department full power 
 to prevent the sewage pollution of all navigable streams and bodies 
 of water under its control, and when all water craft will be com- 
 pelled to dispose of their sewage and garbage in some other man- 
 ner than by dumping it ad libitum into these rivers and lakes, 
 which constitute the sources of water supply for so many populous 
 cities and towns? 
 
 LOWELL, MASS. 
 
 During the autumn and winter of 1890-1891, the city of 
 Lowell, Mass., was visited with a severe epidemic of typhoid fever, 
 which had its origin in two or three cases of typhoid fever in a 
 small manufacturing village (North Chelmsford) situated on Stony 
 Brook, a small stream which flowed into the Merrimac River 
 about three miles above the Lowell Water- Works intake. Within 
 due time after typhoid developed at Lowell, the city of Lawrence, 
 on the Merrimac River about nine miles below Lowell, was also 
 stricken with typhoid fever. The city of Lowell discharges its 
 sewage into the river below the city, while the city of Lawrence 
 below draws its water supply from the river after it has been 
 polluted by the Lowell sewage. 
 
 Thus from Stony Brook the disease was traced to Lowell, and 
 from Lowell to Lawrence. Stony Brook was inoculated through 
 the use of wooden privies overhanging and discharging the fecal 
 matter into the stream, which in turn carried the bacillus of typhoid 
 into the Merrimac River, from which it was pumped into the Lowell 
 
CITATIONS ON TYPHOID FEVER EPIDEMICS. 99 
 
 reservoir, and distributed to the citizens through the public water 
 mains. The fecal discharges of the typhoid patients at Lowell 
 went into the sewers ; these in turn inoculated the river below that 
 city, and furnished the infection to the sister city of Lawrence, 
 which pumped the typhoid-infected water to its reservoir, and like 
 Lowell, supplied it to its citizens through the public water mains. 
 
 Very careful investigations, conducted by Professor W. T. 
 Sedgwick, biologist to the State Board of Health of Massachusetts, 
 clearly established the origin of the epidemic among the people 
 living on Stony Brook, and its carriage to the cities of Lowell and 
 Lawrence through the medium of the public water supply.* 
 
 SAULT STE. MARIE, MICH. 
 
 " The city of Sault Ste. Marie takes its water supply above the rapids, 
 and ordinarily it is very pure water. In the fall of 1890, owing to a break in 
 the lock, more than two hundred boats were detained for two days above the 
 lock and in the immediate vicinity of the intake. Some ten days later typhoid 
 fever appeared, and the deaths were reported to be not less than thirty. All 
 the cases appeared within a few weeks, and sample (analysis below) was taken 
 at this time." f 
 
 ANALYSIS OF WATER FROM SAULT STE. MARIE. 
 
 SOURCE. LAKE MICHIGAN. 
 
 PARTS PER 100,000. 
 
 BACTERIA. 
 
 PER C. C. 
 
 KIND OF 
 BACTERIA FOUND. 
 
 Free Ammonia. 
 
 Albuminoid. 
 
 Chlorine. 
 
 0.0224 
 
 0.0168 
 
 0.33 
 
 2,000 
 
 B. Venensous. 
 
 This water was regarded as the cause of typhoid fever ; and 
 b. venenosus, according to Dr. Vaughan, is classed as a pathogenic 
 organism. " One drachm of a bouillon culture, twenty-four hours 
 old, injected into the abdominal cavity of a rat, produced death." 
 
 ST. Louis, Mo. 
 
 During the year 1892-1893, the city of St. Louis, Mo., U.S.A., 
 had an unusual case and death rate from typhoid fever, the annual 
 mortality from this disease rising from 34 to 103 per 100,000 of 
 population living. 
 
 * Report Massachusetts State Board of Health, 1892, p. 668. 
 
 t A Bacteriological Study of Drinking Water, by Victor C. Vaughan, Ph.D., M.D., Ann 
 Arbor, Mich., 1892, pp. 8-12. 
 
100 THE PURIFICATION OF WATER. 
 
 A careful investigation by the public health department re- 
 vealed the fact that in those districts of the city where the water 
 from the public mains was not generally used for drinking-pur- 
 poses, the citizens were almost wholly exempt from the infection, 
 and that with few exceptions the cases were all confined to a dis- 
 trict the people of which drew all or part of their drinking-water 
 from the public mains.* The water supply of St. Louis is obtained 
 by pumping from the Mississippi River, and the only purification 
 attempted is to allow 24 to 36 hours for sedimentation in several 
 low-level reservoirs. 
 
 ELMIRA, N.Y. 
 
 This city was visited by an epidemic of typhoid fever during 
 January and February of 1896, the cause of which was carefully 
 inquired into by Professor Olin H. Landreth of Schenectady, N.Y. 
 From his very complete report the following facts are gleaned : 
 
 Elmira is situated on the Chemung River, with a population at 
 the time of the epidemic of 37,500. In the cities and villages 
 above Elmira, on the river and its tributaries, and within a reach 
 of 50 miles, an aggregate population of 40,000 is found, among less 
 than three-fourths of which 70 cases of typhoid fever were reported 
 to have occurred prior to the outbreak or increase in the typhoid 
 rates at Elmira. 
 
 Of this number six cases were on premises which sewered to 
 the river, and " a larger number with less definite but equally 
 direct connection with the river." According to the report, the 
 time required for the passage of sewage from the most remote 
 point, where previous cases of typhoid had been recorded, to the 
 water-works intake at Elmira, would be but a few hours. 
 
 During December (1895), there had been fifteen days of con- 
 tinuous cold weather, followed late in the month by heavy rains 
 which caused a freshet in the river, and carried down with the run- 
 off of rainfall considerable polluting matter collected on the 
 frozen ground. 
 
 The epidemic of typhoid in Elmira followed the flood in the 
 Chemung River. 
 
 * Annual Report of Health Commissioner, 1892-93, p. 120. 
 
CITATIONS ON TYPHOID FEVER EPIDEMICS. 101 
 
 ~~ f 
 
 Samples of water from the river at the intake of the Elmira 
 Water- Works, sixteen miles below the nearest .point of sewage pol- 
 lution, taken Dec. 28, 1895, and Jan. 21, 1896, gave from 45,000 
 to 1,080 bacteria per cubic centimeter, among whicJi was found tJie 
 colon bacillus. The larger number of bacteria per cubic centimeter 
 of water was found while the river was in flood or directly after ; a 
 condition often noticed in connection with rivers subject to a wide 
 range of flow. Aside from the very large number of bacteria 
 found at the earlier date of examination, there were other unmis- 
 takable evidences of sewage pollution of the water. 
 
 Search for the typhoid bacillus failed to reveal its presence in 
 the water ; but this is not surprising, for we fail to find it even 
 under much more favorable conditions, and the failure to find it 
 is only negative evidence that it is not in the water. 
 
 After a careful review of all the facts bearing upon the epi- 
 demic, Professor Landreth reached the following conclusions : 
 
 1. The milk supply was in no way responsible for the 
 
 epidemic. 
 
 2. The city water supply was unquestionably the means by 
 
 which the infection was introduced and the epidemic 
 originated. 
 
 3. The city water supply was, in all probability, infected by 
 
 the typhoid bacillus from the Chemung River from cases 
 occurring up-stream during the summer or fall of 1895. 
 
 4. The city water supply as now taken from the collecting 
 
 gallery is quite probably liable to infection at any time 
 from the occurrence of typhoid fever cases near the 
 gallery, in unsewered premises, through the medium of 
 the ground water, and may already have suffered such 
 infection. 
 
 5. A number of cases of fever during the latter stages of the 
 
 epidemic are due to secondary infection from adjacent 
 previous cases, largely through the medium of wells 
 infected from vaults and cesspools. 
 
 6. Well water should be very thoroughly boiled before being 
 
 considered safe for use. It is now more liable to infec- 
 tion than river water. 
 
102 THE PURIFICATION OF WATER. 
 
 PARIS, FRANCE. 
 
 In Popular Science for December, 1895, Mr. Stoddard Dewey 
 is authority for the following statement : 
 
 " In the French army stationed at Paris in 1888, there were 824 cases of 
 typhoid fever, and in 1889, 1,179 cases of typhoid fever. During this time the 
 army had been drinking the sewage-polluted water of the River Seine. In 
 1889 the water of the River Vanne was substituted for that of the Seine, 
 when the number of cases for the next four years, 1890-1893 inclusive, was 
 reduced to 299, 276, 293, and 258. Through an accident the water of the 
 Vanne became contaminated, and for the next three months the cases rose to 
 436. The Vanne again became comparatively free from contamination; and 
 for the next four months of 1895, but eight cases in all occurred, and these 
 were charged to some other water than that of the River Vanne." 
 
 SAN FRANCISCO, CAL. 
 
 The investigation of the public water supply of this city in its 
 relation to typhoid fever, by Dr. M. J. Rosenau, has been men- 
 tioned in a previous chapter ; and in order to appreciate the full 
 force of the facts disclosed, the following brief description of the 
 sources of supply is quoted from the official report to the Super- 
 vising Surgeon-General, U. S. M. H. S. 
 
 " The Spring Valley Water Company furnishes San Francisco most of its 
 water. Its principal source is from three catch basins, made by damming 
 conveniently shaped valleys, between 12 and 20 miles from the city, in San 
 Mateo County. These three reservoirs are known, respectively, as Crystal 
 Springs, San Andreas, and Pilarcitos. From each a pipe-line leads to the 
 city, thus making three water districts. These districts are irregularly shaped, 
 and discontinuous, owing to engineering difficulties. These three systems are 
 connected in such a manner that the water from any one reservoir may be 
 pumped to either of the other two districts. This fact negatives the impor- 
 tance of regarding the districts as separate, from a sanitary standpoint. Smaller 
 reservoirs in the city are used more for the purpose of storage in case of sud- 
 den increased demand, by fire, or from accident, than for obtaining pressure. 
 The flow is continuous. 
 
 "The Visitacion Water Company supplies about 600 consumers in the 
 southern section of the city. The source of this water is from eight wells sunk 
 in Bay View Valley, on the outskirts of the city. Six of these wells are 150 feet 
 deep, one is 180, the other 130 feet through sand and gravel. The wells do not 
 flow spontaneously, and are fitted with deep-well pumps. They are known as 
 
CITATIONS ON TYPHOID FEVER EPIDEMICS. 103 
 
 artesian wells ; but as they do not pass through an imp'ervious stratum, they 
 do not comply with the original meaning of that term. 
 
 " IVells. From the best information at my command, there are very few 
 private wells for drinking and household purposes in the city." 
 
 The usual methods for collecting and planting water samples 
 were followed, the samples b'eing taken from taps in houses. All 
 water was collected by Dr. Rosenau, with the usual precautions to 
 insure a proper sample, and to avoid the introduction of adventi- 
 tious organisms. " Special care was exercised to select a tap that 
 led directly from the mains." Fermenting organisms were iso- 
 lated by the fermentation tube, according to the method proposed 
 by Dr. Theobald Smith. 
 
 " In several instances the colon bacillus was isolated by planting \ cubic 
 centimeter of the suspected water directly upon Wurtz milk-sugar litmus agar, 
 and growing the plants at 41.5-42 C. This slight modification of Smith's 
 method has proven useful, not only for isolating suspicious organisms, but it 
 is believed it may be used to give a proximate idea of their number to the 
 cubic centimeter. 
 
 " In case the above methods failed to find fermenting organisms, the 
 water was further studied upon gelatin and glycerin agar, with the result 
 that, in seven samples, eight suspicious organisms were found. These all 
 proved to be forms belonging to the group of proteus, and resembling b. pro- 
 teus vulgaris. They all grew well at room temperature, also at 37, but not 
 at 41.5 C." 
 
 Thirty-six samples of water collected from widely separated 
 points were examined between Dec. 4, 1895, and Feb. 7, 1896. 
 
 " Fermenting organisms were isolated from 14 of the 36 samples ; 8 of 
 these 14 contained proteus, and the remaining 6 the colon bacillus. ... In the 
 San Andreas water, both the colon bacillus and proteus were found. In 
 the Pilarcitos water also, both proteus and the colon organisms were isolated. 
 In the Crystal Springs water, only an occasional proteus was met with. The 
 greater freedom from fermenting organisms enjoyed by this water is in accord 
 with the control of the watershed, and indicates what may be accomplished 
 in preventing contamination of the water by excluding all habitation from the 
 water basin. The Visitation (deep well) water was found to contain the 
 colon bacillus" 
 
 Concerning tbe colon bacillus, the Report quotes a paragraph 
 from Dr. Abbott * which is worthy of reproduction here. 
 
 * The Principles of Bacteriology, by A. C. Abbott, M.D., Philadelphia, 1894, p. 422. 
 
104 THE PURIFICATION OF WATER. 
 
 " In the normal intestinal tract of all human beings and many other mam- 
 mals, as well as associated with the specific disease producing bacterium in 
 the intestines of typhoid fever patients, is an organism that is frequently found 
 in polluted drinking-waters, and whose presence is proof positive of pollution 
 by either normal or diseased intestinal contents ; and though efforts may result 
 in failure to detect the specific bacillus of .typhoid fever, the finding of the 
 other organism, the bacterium coli communis, justifies one in expressing the 
 opinion that the water under consideration has been polluted by intestinal 
 evacuations from either human beings or animals. Waters so located as to be 
 liable to such pollution can never be considered as other than a continuous 
 source of danger to those using them." 
 
 To which Dr. Rosenau adds, " This organism was found in 
 samples from San Andreas, Pilarcitos, and Visitacion." 
 
 " The fact that the colon bacillus was more readily isolated from the 
 water after the heavy rains, and that fermentation was also more frequent 
 after the heavy rains than before, would indicate soil-washings as one of the 
 sources of this organism in the water ; but that its presence in the water is not 
 alone the result of soil-washings of pasture lands is indicated by a study of the 
 typhoid fever records. Typhoid fever is a constant factor in San Francisco, 
 and a study of (the relation of) the death rate from this disease to (the) rainfall 
 shows a striking coincidence in several years ; i.e., a marked rise in deaths 
 following the first heavy rains." 
 
 In a table covering the period July, 1882, to June, 1895, inclu- 
 sive, Dr. Rosenau has given in parallel columns the total monthly 
 rainfall and deaths from typhoid fever, from which the diagram 
 on the opposite page has been drawn. 
 
 Referring to the diagram, which gives the totals of rainfall and 
 deaths from typhoid fever for fourteen years, it does not appear 
 that the higher typhoid rates are always coincident with the higher 
 rates of rainfall. In fact, the death rate fourth in magnitude for 
 this interval of time occurred during the month of least rainfall 
 (August). Generally, however, the curve of typhoid fever deaths 
 declines with the curve of rainfall, and rises before the period of 
 high precipitation begins. 
 
 DENVER, COL. 
 
 During the past year (1896), the typhoid rates being higher 
 than usual, an investigation of the probable causes was instituted 
 
CITATIONS ON TYPHOID FEVER EPIDEMICS. 
 
 105 
 
 by the health commissioner, with the result that the public water 
 supply was held accountable for the increase of the case and death 
 rates for this disease.* 
 
 Inches of Rainfall for 14 Years. 
 By Months. 
 
 Fig. 2. Totals of Rainfall and Deaths from Typhoia Fever, for a Period of 14 Years, San 
 Francisco, Cat., from January, 1882, to December, 1895. 
 
 (From Notes in Dr. Rosenarfs Report on City Water Supply.) 
 
 The Report contains a table of the deaths from typhoid fever 
 for the years 1892 to 1896 inclusive, for the months of June to 
 
 * Preliminary Report of the Health Commissioner to the Mayor, Denver, Oct. 12, 1896. 
 
106 THE PURIFICATION OF WATER. 
 
 September, from which the months of July, August, and Septem- 
 ber, are reproduced below : 
 
 DEATHS FROM TYPHOID FEVER. 
 YEAR, 1892. 1893. 1894. 1895. 1896. 
 
 July, 28326 
 
 August, 12 4 8 5 13 
 
 September, 9 5 8 8 28 
 
 Total, 23 17 19 15 47 
 
 Commenting upon this the health commissioner says : 
 
 " Is it to be wondered at that such an increase of typhoid fever prevalence 
 in a few weeks should have called attention to the water supply? Investiga- 
 tion of the location of the cases showed that they were fairly well distributed 
 throughout the city; and so far as could.be determined there was no local 
 cause operative in any part of the city to cause this disease, nor was there 
 any one possible cause of infection common to all those affected except the 
 water they drank. . . ." 
 
 " We know that when a certain small percentage of cases due to local in- 
 fection of milk or other food, or contracted by nurses, is eliminated, the great 
 bulk of cases of typhoid fever in all epidemics is due to the one common cause, 
 an infected water supply. When the probable sources of that infection are 
 discovered, it is the plain duty of those upon whom the responsibility lies to 
 suggest the remedies. Failure to adopt the remedial measures that are neces- 
 sary, means deliberate acceptance of responsibility for the resultant loss of 
 human life." 
 
 The city of Denver has two distinct sources of water supply, 
 one from the Platte River, and the other from an artificial reser- 
 voir called Marston Lake, the latter source being fed from moun- 
 tain streams. Both sources were found by the health officials to 
 be exposed to sewage pollution ; and although it was claimed that 
 the water of the river and lake was filtered before it was delivered 
 to the consumers, the Report indicates that this was wholly inef- 
 fectual, and not calculated to improve its quality. 
 
 MlDDLETOWN, CONN. 
 
 This was an epidemic of typhoid caused by oysters which had 
 been fattened in Quinepiac River, near the mouth of a private 
 drain or sewer, from premises where, during the time the oys- 
 
CITATIONS ON TYPHOID FEVER EPIDEMICS. 107 
 
 ters were planted, two cases of typhoid fever Mad occurred. The 
 epidemic followed a banquet or supper given in the autumn of 
 1894 by two of the societies of Wesleyan College, at which some 
 of the students indulged in the oysters raw, while the others ate 
 only cooked oysters. The students who ate raw oysters were the 
 only ones affected. 
 
 The oysters had been taken from distant beds, and replanted 
 in this little stream for the purpose of " plumping" them by hav- 
 ing a current of fresh water run over them during ebb tides. The 
 current of fresh water in this case was probably charged with the 
 germs of typhoid fever from the two cases mentioned above. 
 The reports on the epidemic indicate that the oysters had not 
 been planted at the mouth of the sewer for any great length of 
 time before they were taken up for sale ; but, considering the oys- 
 ter as a natural scavenger of the organic matter in sewage, this 
 time was altogether sufficient for the absorption of some of the 
 sewage and disease germs from the sewer which discharged into 
 the river just above the oyster-bed. 
 
 The students who ate only cooked oysters escaped infection, 
 the cooking being sufficient to sterilize the oysters and such liquid 
 as was contained in the shells. 
 
 In this instance the dejections of the original typhoid patients 
 were thrown into the sewer, and the sewer discharged into the 
 river about three hundred feet above the oyster-bed. The typhoid 
 germs were absorbed by the oysters, and the oysters were ab- 
 sorbed by the students, with the usual result in such cases, that 
 typhoid fever reappeared among the students who ate the unster- 
 ilized oysters. 
 
 STAMFORD, CONN. 
 
 ; ' There has been considerable alarm at Stamford, Conn., over an epidemic 
 of typhoid fever. Over 160 cases appeared in a short time, and the investiga- 
 tion of the State Board of Health showed that all were on the route of one 
 milk-seller. The investigators came to the conclusion that impure milk was 
 the source of the disease. The milkman bought his milk from farmers about 
 Stamford ; but as these farmers also sold their milk to other persons who were 
 not reported ill, it was evident that the germs entered the milk after it came 
 into the possession of the retailer. 
 
108 THE PURIFICATION OF WATER. 
 
 " It was found that the milkman washed his cans in water from a well on 
 his premises, which, upon being analyzed, proved to be totally unfitted for 
 drinking-purposes, and dangerous to use. This case, like the Montclair epi- 
 demic, . . . shows that an infected milk supply may be answerable for many 
 typhoid outbreaks which are not chargeable to a water supply." * 
 
 The mass of evidence on the cause of typhoid fever abroad and 
 in this country is to the effect that it is a water-carried disease ; 
 and even in those epidemics where (as at Stamford, Conn., April, 
 1895, and Montclair, N.J., April, 1894) milk was the immediate 
 distributer of the germ, water used in connection with the dairy 
 operations has been the carrier of the typhoid bacillus. From 
 which it follows, that if the typhoid organism is kept out of our 
 water supplies, typhoid fever would cease to exist as a scourge of 
 the youth and promise of our land.f 
 
 ELIZABETH, N. J. 
 
 " Typhoid fever has become epidemic here, over fifty cases having been 
 reported; and a joint investigation by the Local and State Boards of Health 
 is probable. To the use of polluted water is attributed the outbreak." J 
 
 " There is a marked increase in the number of typhoid fever cases in 
 this city. Not less than forty cases are now being treated by the doctors, 
 and four deaths have occurred, all the victims being adults." 
 
 " E M , aged thirteen, died to-day of typhoid fever at the Alexian 
 
 Hospital. It is the fifth fatal case during the epidemic. His home was on 
 Rahway Avenue, where the disease prevails, and where polluted wells were 
 discovered. Nine cases are in the Elizabeth General Hospital, and House 
 Physician Whitehead is ill, with symptoms of typhoid." || 
 
 EVANSVILLE, IND. 
 
 Dr. Metcalf, of the Indiana State Board of Health, has been in Evans- 
 ville to examine into the cause for the prevalence of typhoid fever. There are 
 one hundred cases of this disease reported by the local physicians. Dr. Met- 
 calf accepted the opinion shared by the physicians that impure water, more 
 than anything else, was at fault. Two of the largest sewers empty into the 
 Ohio River, from which the water supply is taken near the city water-works." ** 
 
 * Engineering Record, May 11, 1895. New York Times, Aug. 21, 1894. 
 
 f Fire and Water, Sept. 22, 1894. || New York Times, Sept. 2, 1894. 
 
 \ New York Times, Aug. 20, 1894. ** Fire and Water, May 11, 1895. 
 
CITATIONS OF TYPHOID FEVER EPIDEMICS. 109 
 
 Thirty-five thousand deaths a year in the 'large cities of the 
 United States are said to be due to typhoid fever alone, a disease 
 the causes of which are fully understood, and which sanitarians 
 declare is entirely preventable. The mischief-making germ is 
 usually taken into the system in drinking-water, which has been 
 contaminated with dejecta from other victims residing many miles 
 away. In some cases when there is a typhoid epidemic, the out- 
 break is traceable to the use of water from a well located near cess- 
 pools ; in others to the common supply of a town, either a river or 
 a lake. To guard against pollution, cities should obtain control of 
 the land around the source of their water supply, and by rigid reg- 
 ulations insure such purity as sensibly to reduce the death rate. 
 But where such a stream as the Mississippi, or such a lake as Michi- 
 gan, is the reliance of any great center of population, precautions 
 of this sort are obviously impracticable. Hence, measures must be 
 taken either by individual consumers or municipal authorities to 
 sterilize the water supply whenever there is occasion for suspicion.* 
 
 The difficulty of impressing upon some people the dangers of 
 a polluted water supply is due to the fact that the drinking of such 
 water is not followed by instant death, j But where is the sub- 
 stantial difference between death from an instant cause, and after 
 two or three or more weeks of wasting fever? The former would 
 be preferred by most men. Because the drinking of a polluted 
 water does not kill at once, like a dose of active poison, it is none 
 the less a poison to some susceptible systems ; and any corporation 
 which, after the fact of pollution is known, refuses to supply a safe 
 water when a safe water can be had, or refuses to so deal with the 
 present water as to render it innocuous to health, is deserving of 
 the just censure of a suffering public. :f 
 
 The following quotation would seem to indicate that this view 
 of the matter is sometimes taken : 
 
 " Warren, Ohio, won a great victory over the Warren Water Company. 
 The judgment for $7,621.35, obtained by the company against the city in the 
 lower court recently, was canceled by the Circuit Court, which held that the 
 company had not furnished pure water, as bound to do under its contract." 
 
 * New York Weekly Tribune, Oct. 3, 1894. \ See Appendix C. 
 
 f Fire and Water, Sept. 22, 1894. New York Tribune, Nov. 18, 1894. 
 
110 THE PURIFICATION OF WATER. 
 
 CHAPTER VIII. 
 
 SEDIMENTATION OF POLLUTED WATERS. 
 
 EXPERIMENTAL information upon the influence of sedimentation 
 on the quality of water supplies is rather meager ; although the fact 
 is well known that sedimentation for even a few hours, with certain 
 waters, has a marked feect on their appearance, while sedimenta- 
 tion for great lengths of time, according to Miquel, has wholly 
 eliminated the evidences of organic matter and bacteria in the 
 water of the River Seine. It is reasonable to suppose that in due 
 time all organic matter in a polluted water, contained in a subsid- 
 ing or impounding reservoir, would be appropriated by the bacteria 
 of putrefaction and converted into ammonia compounds ; and these, 
 then acted upon by the nitrifying organisms, and converted into 
 nitrous and nitric acids, which uniting with the bases, such as lime, 
 magnesia, soda, potash, etc., in the water, will be precipitated as 
 insoluble harmless compounds. 
 
 Such sedimentation as is really effective in connection with 
 works of public water supply is probably limited to the precipita- 
 tion by gravity alone of the suspended matters which give color 
 to the water, such as sand, clay, and the complex combination of 
 inorganic and organic matter called "silt" Some chemical and 
 biologic changes occur in the water during this precipitation of 
 suspended matter by gravity ; but when the sedimentation is 
 limited to a few hours or a few days, it is doubtful if any marked 
 improvement in the quality has taken place. 
 
 Some improvement in color of turbid waters will usually take 
 place within a few hours, unless the color is due to finely divided 
 peaty substances, but no change in the organic matter or bacterial 
 contents and species will in most instances occur. 
 
 Sedimentation of a character which will really affect^ the hygiene 
 of water requires great length of time, and should be conducted in 
 
SEDIMENTATION OF POLLUTED WATERS. 
 
 Ill 
 
 very large deep reservoirs, which precludes an attempt at purifi- 
 cation of polluted waters by sedimentation alone by most cities. 
 
 The following table contains the counts and percentages of re- 
 duction of bacteria, in the water of the Ohio River, for an interval 
 of time not exceeding thirty-two days. The Cincinnati tap water, 
 although pumped to a distributing reservoir, really has no time for 
 sedimentation before it leaves this reservoir and is used by the 
 consumers ; while the Covington, Ky., reservoirs are of such ca- 
 pacity in relation to the daily consumption, that usually there is 
 about one month's time allowed for subsidence before the water 
 passes from the reservoirs to the consumers. 
 
 REDUCTION OF BACTERIA BY SUBSIDENCE.* 
 
 DATE OF 
 INOCULATION. 
 
 DAYS OF 
 GROWTH ON 
 GELATIN. 
 
 BACTERIA 
 PER C. C. OF WATER. 
 
 PERCENTAGE 
 OF BACTERIA 
 IN COVINGTON 
 WATER. 
 
 REDUCTION BY 
 SEDIMENTATION. 
 
 PER CENT. 
 
 Cincinnati. 
 
 Covington. 
 
 1896 
 Jan. 17, 
 
 5 
 
 1,472 
 
 272 
 
 18.50 
 
 81.50 
 
 " 23, 
 
 4 
 
 1,599 
 
 194 
 
 12.13 
 
 87.87 
 
 " 28, 
 
 4 
 
 5,002 
 
 172 
 
 3.39 
 
 96.61 
 
 " 28, 
 
 4| 
 
 
 182 
 
 3.59 
 
 96.41 
 
 Feb. 4, 
 
 4| 
 
 1,656 
 
 53 
 
 3.20 
 
 96.80 
 
 " 4, 
 
 6 
 
 2,042 
 
 56 
 
 2.74 
 
 97.26 
 
 " 8, 
 
 n 
 
 1,561 
 
 63 
 
 4.04 
 
 95.96 
 
 " 11, 
 
 *i 
 
 1,526 
 
 75 
 
 4.95 
 
 95.05 
 
 " 17, 
 
 7 
 
 684 
 
 20 
 
 2.92 
 
 97.08 
 
 " 21, 
 
 4 
 
 329 
 
 26 
 
 7.90 
 
 92.10 
 
 " 21, 
 
 7 
 
 1,232 
 
 112 
 
 9.09 
 
 90.91 
 
 " 26, 
 
 3} 
 
 1,144 
 
 84 
 
 7.34 
 
 92.66 
 
 " 26, 
 
 5 
 
 1,436 
 
 102 
 
 7.10 
 
 92.90 
 
 The numbers of bacteria and percentage of reduction in the 
 Covington water were at times very gratifying, and indicate what 
 may be expected in situations where the water can be carried in a 
 quiescent state in large reservoirs for several months before it is 
 drawn off for use. 
 
 According to Professor Percy Frankland, the average number 
 of bacteria in the two streams which discharge into the Loch Lin- 
 thrathen, the source of water supply of Dundee, was 1,240, while 
 
 * Report of Engineer Commission on Extension and Betterment of Cincinnati Water 
 Works, 1896, p. 15. 
 
112 
 
 THE PURIFICATION OF WATER. 
 
 the average number of bacteria in the water issuing from the lake 
 was 30, showing a reduction in bacterial contents of the water by 
 subsidence of 97.6 per cent. 
 
 The same authority gives the following results from samples of 
 water from the West Middlesex Works for 1892 : 
 
 BACTERIA PER C. C. OF WATER. 
 
 Thames water at Hampton, 1,437 
 
 Same water after passing two storage reservoirs, 177 
 
 Showing a reduction of 87.7 per cent in the bacteria by a few days' 
 subsidence. 
 
 In the report of bacterial examinations of the river waters 
 supplied to London for 1895, Dr. E. Frankland gives the following 
 data : 
 
 AVERAGE FOR TWELVE MONTHS. 
 
 SOURCE OF WATER. 
 
 BACTERIA PER 
 C. C. 
 
 PERCENTAGE 
 OF REDUCTION. 
 
 Thames at Hampton, 
 
 13,646 
 
 . . . 
 
 Chelsea reservoirs, 13 days' storage, 
 
 3,177 
 
 76.7 
 
 West Middlesex reservoir, 6.3 days' storage, 
 
 971 
 
 92.9 
 
 Lambeth reservoir, 6.4 days' storage, 
 
 3,520 
 
 74.0 
 
 Grand Junction reservoir, short storage, 
 
 917 
 
 94.0 
 
 River Lea, Angel Road, 
 
 14,075 
 
 . . . 
 
 East London reservoirs, 15 days' storage, 
 
 6,280 
 
 55.4 
 
 The reductions above are obtained from natural subsidence, 
 without the aid of chemicals. The author's experiments with 
 potash alum and slaked lime, on the suspended matter in the 
 Ohio River water have given the following results : 
 
 TAP WATER TREATED WITH POTASH ALUM AND SLAKED LIME; SUSPENDED 
 MATTER ALLOWED TO SUBSIDE DURING 24 HOURS IN ICE-CHEST. 
 
 DATE. 
 
 KIND OF WATER. 
 
 DAYS OF 
 GROWTH ON 
 GELATIN. 
 
 COLONIES 
 
 PER C. C. 
 
 OF WATER. 
 
 Dec. 
 <( 
 <( 
 
 11, 1896, 
 
 a 
 
 (i 
 
 Plain tap water, 
 Treated with 2.57 gr. of alum per gal., 
 Treated with 3. 74 gr. of slaked lime per gal. , 
 
 4 
 3 
 4 
 
 11,021 
 1,674-1,803 
 55- 59 
 
 Reduction by alum (without filtration) per cent, 84.22 
 Reduction by lime (without filtration) per cent, 99.48 
 
SEDIMENTATION OF POLLUTED WATERS. 113 
 
 Another test of the same water, treated with two and a half 
 milligrams of potash alum to two ounces of water, kept in ice-chest 
 for forty-eight hours, sample taken without disturbing sediment in 
 bottle, cultivated on gelatin, gave the following results : 
 
 DATE. DAYS OF GROWTH. COLONIES PER C. C. OF WATER. 
 
 December 13, 1896, 4 1,866-2,315 
 
 Reduction by alum (without filtration) per cent, 81.04 
 
 Hard water from the Colne Valley Water- Works, according to 
 Professor Percy Frankland, contains 322 bacteria per cubio centi- 
 meter, and after treatment for reduction of hardness by the Clark 
 process, with two days' subsidence, contains 4 bacteria per cubic 
 centimeter, indicating a reduction of the bacterial contents, by the 
 lime process and two days' sedimentation, of nearly 99 per cent. 
 
 Another experiment by the same authority, using the Clark pro- 
 cess in combination with a mechanical separator in which 4 only two 
 hours were allowed for deposition of the lime, gave the following 
 results : 
 
 Artesian well water (London) contained 182 bacteria per cubic 
 centimeter, while the treated water issuing from the mechanical 
 separator contained 4 bacteria per cubic centimeter, showing a 
 reduction of nearly 98 per cent in the bacterial contents of the 
 water. 
 
 The following experiments upon assisted subsidence of the 
 bacteria in water are given by Professor E. Ray Lankester, in his 
 evidence before the Royal Commission on Metropolitan Water 
 Supply : * 
 
 EXPERIMENTS SHOWING EFFECT OF SUBSIDENCE OF MUD AND CLAY ON 
 THE NUMBER OF BACTERIA IN SUSPENSION IN WATER. 
 
 1. Three jars, A, B, C, each holding one liter of Oxford tap 
 water, were taken on June 27, 1892. To A were added and well 
 stirred in, 25 grams of sterilized kaolin ; to B, similarly 12 grams 
 of sterilized kaolin ; C was untouched. After 15 hours the kaolin 
 had completely subsided, and plate cultures were made from each 
 
 * Appendix C, p. 456. 
 
114 THE PURIFICATION OF WATER. 
 
 jar in order to determine the relative number of bacteria now in 
 suspension in the water. 
 
 From A, 1,200 colonies per c. c. were obtained. 
 " B, 2,790 
 " C, 7,040 
 
 Repetitions of the experiment yielded similar results, showing 
 that as much as five-sixths of the bacteria present were carried down 
 by ; the subsiding kaolin, when added in proper quantity. 
 
 N 2. On July 7 river mud sterilized by heat (80 degrees C.) was 
 substituted for kaolin. To one liter jar of Oxford tap water 30 
 grams of the river mud were added and stirred in, while a second 
 liter jar of the same water was kept for comparison. The jars 
 stood undisturbed for 20 hours. The river mud having now com- 
 pletely subsided, plate cultures of the water in the two jars were 
 made. That to which nothing had been added showed 55,000 
 colonies per cubic centimeter, while that to which the river mud 
 had been added showed only 15,400 colonies. 
 
 The experiment with sterilized river mud was repeated on July 
 26 and 27 with similar results ; viz., a reduction of the bacteria by 
 subsidence of the mud to the extent of nearly three-fourths. 
 
 (These experiments have importance not only for the history of 
 bacteria in the river normally, but especially for the question of 
 the storage of flood water. The subsidence of the mud suspended 
 in such flood water would largely tend to purify the water from 
 any excess of bacteria.) 
 
 3. Experiments on addition of lime to river water. June 18, 
 1892, two liter jars of river water (Thames, Oxford) were taken ; 
 to one, 6 grams of slaked lime were added. After two days, plate 
 cultivations were made from the water in each jar ; that to which 
 no lime had been added showed 5,000 colonies ; that to which the 
 lime had been added showed only 280 colonies per cubic centimeter. 
 
 Reduction by lime 94.4 per cent. 
 
 A similar experiment with tap water on June 20 gave 120 
 colonies without lime as against 15 colonies where the lime had 
 been added. 
 
 Reduction by lime 87.50 per cent. 
 
SEDIMENTATION OF POLLUTED WATERS. 
 
 116 
 
 Alum has a still more remarkable effect than lime on the 
 bacteria in river water. On the 17th of September two liters of 
 tap water were taken for comparison. To one (liter) \ gram of 
 alum was added. After subsidence (i.e., 24 hours) the untouched 
 water gave (1) 15,130 colonies, while the water to which the alum 
 had been added gave none at all. On the 26th of September a 
 similar experiment gave (2) 2,380 colonies in the untouched water, 
 and 8 in that to which alum had been added. 
 
 Reduction by alum (1) = 100.00 per cent. 
 
 Reduction by alum (2) = 99.66 per cent. 
 
 The alum used by Professor Lankester amounted to 14.60 
 grains per gallon, which at 1.6 cents per pound would make the 
 cost of treatment per million U. S. gallons $33.32 for chemicals 
 alone. No mention is made of the presence of undecomposed alum 
 in the water, but it cannot be doubted that with such a proportion 
 of alum a considerable astringency must have been imparted to 
 the water. 
 
 EXPERIMENTS ON THE PRECIPITATION OF THE SUSPENDED MATTER 
 IN OHIO RIVER WATER. 
 
 The data detailed in the table below were collected by Mr. 
 Edward Flad, C.E., in connection with certain experiments on 
 sedimentation for the city of Cincinnati, 1889, and indicate that 
 sedimentation for an interval as short as 40 hours will reduce the 
 suspended matter by weight nearly 80 per cent. 
 
 EXPERIMENTS ON THE PRECIPITATION OF THE SUSPENDED MATTER IN 
 OHIO RIVER WATER AT CINCINNATI, OHIO. 
 
 No. OF 
 SAMPLE. 
 
 DATE, 1889. 
 
 HOURS OF 
 SETTLING. 
 
 SILT HELD IN SUSPENSION 
 PARTS BY WEIGHT PER 1,000. 
 
 PERCENTAGE 
 OF SILT 
 REMOVED BY 
 SETTLING. 
 
 Before Settling. 
 
 After Settling. 
 
 4 
 
 Jan. 7 
 
 42.2 
 
 0.3635 
 
 0.1225 
 
 66.3 
 
 2 
 
 9 
 
 47.0 
 
 0.3610 
 
 0.1135 
 
 68.5 
 
 5 
 
 11 
 
 46.3 
 
 0.2350 
 
 0.1435 
 
 38.9 
 
 7 
 
 13 
 
 48.0 
 
 0.1005 
 
 0.0490 
 
 51.2 
 
 8 
 
 15 
 
 47.1 
 
 0.0920 
 
 0.0330 
 
 64.1 
 
 14 
 
 17 
 
 47.3 
 
 0.3900 
 
 0.1305 
 
 66.5 
 
 13 
 
 19 
 
 46.5 
 
 0.1590 
 
 . . . 
 
 . . . 
 
 17 
 
 21 
 
 48.2 
 
 0.2011 
 
 0.0932 
 
 53.6 
 
 26 
 
 23 
 
 41.5 
 
 0.0865 
 
 0.0246 
 
 71.5 
 
116 
 
 THE PURIFICATION OF WATER. 
 
 EXPERIMENTS ON PRECIPITATION.- Continued. 
 
 No. OF 
 SAMPLE 
 
 DATE, 1889. 
 
 HOURS OF 
 SETTLING. 
 
 SILT HELD IN SUSPENSION 
 PARTS BY WEIGHT PER 1,000. 
 
 PERCENTAGE 
 OF SILT 
 REMOVED BY 
 SETTLING. 
 
 Before Settling. 
 
 After Settling. 
 
 24 
 
 Jan. 25 
 
 30.4 
 
 0.0405 
 
 0.0540 
 
 
 19 
 
 26 
 
 40.4 
 
 0.0955 
 
 0.0220 
 
 75.9 
 
 29 
 
 29 
 
 5.3 
 
 0.1640 
 
 0.0720 
 
 56.1 
 
 30 
 
 29 
 
 40.3 
 
 0.2235 
 
 0.0580 
 
 74.0 
 
 31 
 
 31 
 
 30.3 
 
 0.2225 
 
 0.1098 
 
 50.6 
 
 33 
 
 Feb. 1 
 
 41.2 
 
 0.3095 
 
 0.0720 
 
 76.8 
 
 34 
 
 3 
 
 31.5 
 
 0.2760 
 
 0.0910 
 
 67.0 
 
 38 
 
 4 
 
 42.0 
 
 0.1900 
 
 0.0445 
 
 76.5 
 
 40 
 
 6 
 
 28.2 
 
 0.1615 
 
 0.0615 
 
 61.9 
 
 43 
 
 7 
 
 40.3 
 
 0.1548 
 
 0.0560 
 
 63.8 
 
 61 
 
 9 
 
 30.3 
 
 0.0555 
 
 0.0325 
 
 39.6 
 
 60 
 
 10 
 
 40.4 
 
 0.0450 
 
 0.0220 
 
 51.1 
 
 44 
 
 12* 
 
 30.6 
 
 0.0415 
 
 0.0360 
 
 13.2 
 
 62 
 
 13 
 
 41.1 
 
 0.0462 
 
 0.0188 
 
 69.3 
 
 67 
 
 15 
 
 30.2 
 
 0.0665 
 
 0.0125 
 
 81.2 
 
 69 
 
 16 > 
 
 39.4 
 
 0.2635 
 
 0.0330 
 
 85.8 
 
 77 
 
 18 
 
 31.0 
 
 0.5425 
 
 0.1287 
 
 76.3 
 
 74 
 
 19 
 
 40.5 
 
 0.5900 
 
 0.1085 
 
 81.6 
 
 75 
 
 21 
 
 31.5 
 
 0.5623 
 
 0.1628 
 
 71.1 
 
 78 
 
 22 
 
 41.2 
 
 0.3780 
 
 0.0945 
 
 75.0 
 
 80 
 
 24 
 
 29.6 
 
 0.3455 
 
 0.0855 
 
 75.3 
 
 83 
 
 25 
 
 40.3 
 
 0.3811 
 
 0.0930 
 
 75.6 
 
 84 
 
 27 
 
 47.1 
 
 0.2940 
 
 0.0765 
 
 74.0 
 
 A review of the data in the table indicates that the greatest 
 percentage of reduction of the silt accompanies the greatest tur- 
 bidity of the river water. 
 
 Experiments by the author upon the rate of reduction of the 
 suspended matter in the Ohio River water have given the follow- 
 ing results : 
 
 SEDIMENTATION OF SUSPENDED MATTER IN OHIO RIVER WATER, 
 DECEMBER, 1896. 
 
 PARTS PER 100,000. 
 
 Original amount of matter in suspension, 54.00 
 
 Matter in suspension at end of two days, 15.00 
 
 Matter in suspension at end of four days, 13.50 
 
 . Matter in suspension at end of six days, 12.00 
 
 Showing a reduction of 72 per cent of the suspended matter in 
 
 two days, 75 per cent in four days, and 77.8 per cent in six days. 
 
 The rate of subsidence will depend upon the specific gravity of 
 
SEDIMENTATION OF POLLUTED WATERS. 
 
 117 
 
 the matter in suspension, and the quiescence of the water under- 
 going sedimentation. When the specific gravity of the suspended 
 matter is considerably in excess of one, and the water altogether 
 at a state of rest, the precipitation will be rapid. Conversely, with 
 a specific gravity of the suspended matter not much above that of 
 water, and the water in a state of agitation, the sedimentation will 
 be slow, and under unfavorable conditions there may be no precipi- 
 tation by subsidence at all. 
 
 The reduction of organic matter in water by subsidence is due 
 partly to precipitation of matters heavier than water, and partly to 
 destruction of organic matter in suspension by bacterial action ; 
 while the reduction of the bacterial contents of water by sedimen- 
 tation is accomplished partly by precipitation of the bacteria in 
 contact with the suspended matter, and partly by the natural 
 decay of the less hardy species of the water bacteria. In the 
 biologic action which occurs in large bodies of water, the weaker 
 species of the bacteria, as organic matter, are absorbed by or be- 
 come food for the stronger species ; and this process of the destruc- 
 tion of the weaker by the stronger forms goes on until all food 
 supply is exhausted, whereupon the strongest species 'perish, and 
 in a manner still to be explained are converted into the harmless 
 nitrogenous compounds, and as such are precipitated along with 
 the inorganic matters in suspension in all surface waters. 
 
 REDUCTION OF HARDNESS AND BACTERIAL CONTENTS 
 BY ADDITION OF LIME. 
 
 Experiments conducted by Mr. Dibdin,* on the water of the 
 New River Company furnished the following results : 
 
 DATE. 
 
 BEFORE TREATMENT. 
 
 AFTER TREATMENT. 
 
 PERCENTAGE OF 
 REDUCTION. 
 
 
 Hardness. 
 
 Bacteria. 
 
 Hardness. 
 
 Bacteria. 
 
 Hardness. 
 
 Bacteria. 
 
 Dec. 16, 1895. 
 
 17.4 
 
 96 
 
 5.4 
 
 12 
 
 68.4 
 
 87.5 
 
 Dec. 18, " 
 
 17.4 
 
 110 
 
 5.0 
 
 6 
 
 71.3 
 
 94.5 
 
 Dec. 20, " 
 
 17.4 
 
 60 
 
 8.5 
 
 16 
 
 51.2 
 
 73.4 
 
 1896. 
 
 Analytical Investigations of London Water Supply, London County Council, January, 
 
118 THE PURIFICATION OF WATER. 
 
 The lime treatment reduced the red color in the water, by 
 Lovibond's tintometer, 100 per cent, and the yellow color 35 per 
 cent. No change was noticed in the "free ammonia" after and 
 before treatment, while the albuminoid ammonia was reduced 23 
 per cent. The chlorine was unaffected by the treatment, while the 
 oxygen absorbed on a four-hour test before and after treatment was 
 reduced 25 per cent. 
 
 The total solids were reduced from an average of 24.3 parts 
 per 100,000 parts of water before treatment to 11.8 parts after 
 treatment. In all the tests the water was dosed with 9.4 per cent 
 of a saturated lime-water. Of these tests Mr. Dibdin says, " It 
 would therefore seem that by the adoption of the system of soft- 
 ening, the present supply, in respect to its chemical quality and 
 bacteria, would be improved to a degree comparable with that of 
 the Welsh sources." * 
 
 Such experience as has been had along the line of reduction of 
 hardness in water for city supply has revealed the interesting fact, 
 that the addition of lime to a hard (polluted) water is effective in 
 the purification of the water, as well as in the reduction of the 
 hardness, as indicated by the experiments previously noted. 
 
 Elaborate appliances for the lime treatment of water are now 
 being built by several companies abroad, and the author is in- 
 formed that the Jewell Filter Company of Chicago has built some 
 apparatus for this purpose for cities in this country. 
 
 With reference to the cost of water softening on .a very large 
 scale, the following information is abstracted from Reports by Mr. 
 W. J. Dibdin, chemist, and Sir Alex. R. Binnie, engineer, to the 
 London County Council, on the London Water Supply for 1895. 
 
 Considering, according to Mr. Dibdin, the cost of lime alone 
 for a daily treatment of 200,000,000 imperial gallons, this will 
 amount to 35,000 or $175,000 per year. Considering the cost 
 
 * It is not within the province of this work to discuss projects of water supply ; but it may be 
 of interest to mention that the Welsh sources with which Mr. Dibdin compares the lime-treated 
 London water are ably and elaborately presented in a Report by Sir Alexander R. Binnie to the 
 London County Council, entitled, Available Sources of Water Supply for London, June, 1894. 
 
 The development of these sources in Wales will provide a daily supply of 415,000,000 im- 
 perial gallons, at an estimated cost of nearly $200,000,000, the water to be conducted to the 
 metropolis through two aqueducts 150 and 176 miles long. 
 
SEDIMENTATION OF POLLUTED WATERS. 119 
 
 of lime and all labor, Mr. Binnie puts this at* .300 to 310 per 
 million imperial gallons treated daily per year, while the cost of ap- 
 paratus and buildings for lime treatment, he estimates at .3,500 to 
 4,500 or $20,000 per million imperial gallons treated daily. For 
 the present daily consumption of water by London (200,000,000 
 imperial gallons), Mr. Binnie puts the annual cost at 60,000 or 
 $300,000. 
 
 Reducing these figures to our measures and values, the annual 
 cost for 1,000,000 U. S. gallons treated per day, for a reduction 
 from a hardness of 17 degrees by Clark's scale to about 5 degrees 
 for lime alone, according to Mr. Dibdin, becomes 1708.00 ; and 
 the whole cost of lime and labor, according to Mr. Binnie, will be 
 $1,270.32, or about ^ cent per 1,000 gallons of water treated. 
 These figures are based upon the treatment of 240,000,000 U. S. 
 gallons per day, and this cost would naturally not be applicable to 
 small quantities of water per diem. 
 
120 THE PURIFICATION OF WATER. 
 
 CHAPTER IX. 
 
 STERILIZATION OF DRINKING-WATER.* 
 
 SEVERAL years ago, in discussing the hygiene of public water 
 supply, the author took the ground, that, as the proportion of water 
 used for drinking-purposes was one-half per cent or less of the 
 whole quantity consumed by the takers from a public source, the 
 better plan was not to attempt to secure the whole supply of po- 
 table quality, but to render any water available fit for drinking- 
 purposes by domestic filtration. Later experience satisfies him 
 that this plan will not answer for several reasons : 
 
 1. All consumers of a public water supply cannot, or will not, 
 
 use domestic filters. 
 
 2. There is no domestic filter which is absolutely proof against 
 
 the dangers of polluted water. 
 
 3. Even if a satisfactory filter was obtainable, it is doubtful if 
 
 the average householder would give this the attention 
 it requires to keep it at all times in condition to act as 
 a safeguard. 
 
 In view of which the conclusion has been reached, that if the 
 consumer is to have a safe drinking-water, it must come to him in 
 this condition through the public water mains. In other words, 
 the matter of purity must be looked after by the municipal corpo- 
 ration or the water company. The prevalence of typhoid fever in 
 any city or town having a public water supply is evidence that 
 the water now generally furnished to consumers is unpotable, and 
 that municipal corporations and water companies are delivering to 
 their consumers water containing the specific organism of typhoid 
 fever. 
 
 * A portion of this chapter is abstracted from a paper by the author, read at the Eighth 
 International Congfess of Hygiene and Demography, Buda-Pest, Austria, September, 1894. 
 
STERILIZATION OF DRINKING-WATER. 
 
 It is common for physicians in case of doubt of the purity of a 
 water supply, to recommend that water for drinking-purposes be 
 boiled ; but the boiling of water renders it insipid and unpalatable, 
 and it is claimed by some of the manufacturers of filters that water 
 deprived of certain of its natural gases and solids in solution (as it 
 will be by boiling) is not as wholesome as natural waters. The 
 author has been unable to obtain any reliable information of the 
 influence on the human system of the salts and gases in solution 
 in natural waters, and is uncertain whether the continuous use of 
 boiled water as a beverage will be deleterious. Considering that 
 filtered and boiled water will be limpid and sterile and deprived of 
 all toxic properties, and assuming that such water will not be in- 
 jurious to the system, may not the problem of an absolutely safe 
 drinking-water finally be solved by combined filtration and distilla- 
 tion ? If carried out to its legitimate conclusion this would mean 
 the treatment of a sufficient quantity of water by the municipal 
 corporation for drinking and culinary purposes, and the delivery of 
 this to consumers through an independent system of comparatively 
 small mains. But the expensive apparatus for distillation ; the cost 
 of duplicating the street mains, even with pipes of small diameter ; 
 and especially the large annual expense of operation, might at 
 first sight seem to prohibit any attempt by this process to purify 
 water on a large scale. 
 
 For the purpose of estimating the probable cost of this method 
 of water purification for city use, let us take an American city 
 with a population of 400,000, and allow a daily consumption of 
 water for all purposes of 40,000,000 U. S. gallons, or 100 gallons 
 per head of population ; of which quantity it will be assumed 
 that 2^ per cent, or 1,000,000 gallons per day, is used exclu- 
 sively for drinking and cooking purposes, including water for the 
 washing of culinary vessels and apparatus. To sterilize by heat 
 1,000,000 U. S. gallons of water per day of 24 hours will require 
 an hourly distillation of 347,100 pounds ; and assuming the average 
 temperature of the filtered water (or feed water) to the boilers 
 to be 60 Fahr., and the pressure of distillation to be six pounds 
 above the atmosphere, then the total heat to be added to each 
 pound of water will be 1,124 B. T. u. 
 
122 THE PURIFICATION OF WAITER. 
 
 If the steam in going from the boilers to the surface con- 
 densers be made to pass through suitable closed heaters, through 
 which also the cold water to the boilers is being pumped, then a 
 part of the heat of the steam will be given up to the feed water, 
 and a smaller amount of heat will be required from the coal or 
 other fuel to sterilize a given amount of water, and a smaller ca- 
 pacity of boilers and surface condensers will be required. Since 
 the cold water from the niters will be pumped under full boiler 
 pressure through these closed heaters, it will be possible (if such 
 heaters are of sufficient capacity) to supply to the water not only 
 the sensible heat, but a part of the latent heat from the heat in 
 the steam before it is finally condensed in the surface condensers ; 
 or of the heat in the sterilized water a large percentage can be 
 recovered and utilized in heating the filtered water to the boilers, 
 with a corresponding saving of fuel. 
 
 The cost of fuel being the bete notrin the problem of steriliz- 
 ing by heat the drinking-water for a city, it is desirable that the 
 facts in connection with the expenditure of fuel be carefully and 
 fully considered. The rate at which the feed water is pumped to 
 the boilers being the same as the rate of flow of steam through 
 the heaters to the condensers, it follows, that, if the heaters were 
 large enough and sufficient time allowed for the passage of the 
 steam through them, and there were no losses of heat by radia- 
 tion, etc., one-half of the heat of the steam would be transferred 
 to the water on its way to the boilers ; or of the 1,124 heat units 
 added per pound of water in the boilers, 562 units would be car- 
 ried back in the feed water. But the recovery of 50 per cent of 
 the heat assumes an efficiency beyond the reach of ordinary heat- 
 ing apparatus ; and some allowance must be made for the losses 
 by conduction and radiation, and by contact of air, which can safely 
 be put at 10 per cent ; and considering the very slow rate of 
 transfer of heat when the temperature of the steam (partially con- 
 densed) and that of the feed water approximate each other, it 
 will be safe to allow another 10 per cent loss upon account 
 of time ; from which, as a practical proposition, it is estimated 
 that of the heat carried off from the boilers by the steam, 30 
 per cent may be recovered in the feed-water heaters, leaving 
 
STERILIZATION OF DRINKING-WATER. 123 
 
 70 per cent to be taken up by the cooling water in the surface 
 condensers. 
 
 It will therefore be necessary to supply to each pound of water 
 pumped into the boilers 1,124 x .7 = 787 heat units ; and with 
 coal and boilers showing an efficiency of 11,250 heat units per 
 pound of fuel, each pound of coal will distill 14.3 pounds of water 
 at 6 pounds pressure above the atmosphere ; and for the distilla- 
 tion or sterilization of 347,100 pounds per hour (1,000,000 U. S. 
 gallons per day), there will be required 291 tons (2,000 pounds) 
 of coal, or an annual consumption of 106,215 tons. 
 
 The boiler capacity to distill this amount of water daily has 
 been estimated as follows : An ordinary return tubular boiler sup- 
 plied with water at 60 Fahr., and working at six pounds pressure 
 above atmosphere, will easily evaporate 3.5 pounds per hour per 
 square foot of heating surface ; and if the heat required per pound 
 of water be 787 instead of 1,124 thermal units, then each superfi- 
 cial foot of heating surface can be expected to evaporate 5 pounds 
 of water per hour ; and for the evaporation of 347,100 pounds per 
 hour, there will be required a-u^sm = 69,420 square feet of heat- 
 ing surface. Allowing 2,000 square feet to each boiler, there will 
 be required 35 boilers, each 6 feet 6 inches diameter by 18 feet 
 long, with the proper complement of tubes. The feed-water heat- 
 ers to heat the filtered water and partially condense the steam from 
 the boilers have been estimated in the following manner : 
 
 Each square foot of surface, taking the tube in the heater as 
 -fV inch or less in thickness, will readily transfer 4,000 thermal 
 units per hour, equivalent to the heating through 337 Fahr. of 
 12 pounds of water ; and to heat 3-7o_o pounds will require 
 29,000 square feet of heating (or cooling) surface in the heaters ; 
 or with an allowance of 1,000 square feet of surface to each 
 heater, there will be required 29 heaters to deal with 1,000,000 
 gallons of water per day. 
 
 Surface condensers constructed with thin brass tubes can be 
 estimated to condense 15 pounds of steam per square foot of cool- 
 ing surface per hour ; and if these also are of 1,000 square feet 
 each, there will be required, to deal with 1,000,000 gallons of water 
 in 24 hours, 23 such condensers. 
 
124 THE PURIFICATION OF WATER. 
 
 The apparatus, therefore, which we have outlined for the sterili- 
 zation by heat of 1,000,000 U. S. gallons of water per day, consists 
 in detail of a duplicate filter plant, each half of 1,000,000 gallons 
 daily capacity ; pumping machinery in duplicate of 1,000,000 
 gallons daily capacity to take the filtered water and supply it to 
 the boilers ; steam-boilers to evaporate the water under low pres- 
 sure ; closed feed-water heaters to cool the steam and heat the 
 feed water ; surface condensers to condense the steam ; and pump- 
 ing machinery to take the condensed steam and sterilized water, 
 and pump it into the mains for distribution to the consumers. 
 The filters, heaters, and steam-boilers will require buildings for 
 their protection from the weather, while the condensers may be 
 exposed to the weather without detriment to their operation or 
 durability. In addition to the apparatus mentioned, for a city 
 of the population we have assumed, there will be required about 
 350 miles of mains of small diameter, to distribute the sterilized 
 water to the various premises to be supplied. 
 
 We are now ready to estimate the cost of constructing and 
 operating such a plant for water purification : 
 
 COST OF CONSTRUCTION. 
 
 Two filter plants, each of 1,000,000 gallons daily 
 
 capacity, $15,000.00 
 
 Filter-house, 4,000.00 
 
 Thirty-five steam-boilers, complete, 63,000.00 
 
 Twenty-nine feed-water heaters, complete, 29,000.00 
 
 Twenty-three surface condensers, complete, 34,500.00 
 
 Boiler-house, 30,000.00 
 
 Two sets of pumping machinery, each set of 1,000,000 
 gallons daily capacity, to supply the filtered water 
 to the boilers, 9,000.00 
 
 Two sets of pumping machinery, each set of 1,000,000 
 gallons daily capacity, to pump the sterilized water 
 into the mains, 12,000.00 
 
 Pumping-station, 12,000.00 
 
 Add for pipes, valves, etc., at sterilizing-station, 20,000.00 
 
 350 miles of mains at an average cost of $4,500 per 
 
 mile, 1,575,000.00 
 
 Total, $1,803,500.00 
 Cost per capita of population, $4.50. 
 
STERILIZATION OF DRINKING-WATER- 125 
 
 FIXED ANNUAL CHARGES. 
 
 Interest on cost of construction at 5^>, $90,175.00 
 Annual payment to sinking-fund to redeem construc- 
 tion bonds invested at 4^ for 40 years, 18,972.82 
 Total, $109,147.82 
 
 OPERATING EXPENSES. 
 
 106,215 tons of coal at $2, $212,430.00 
 
 Forty-five men at $2 per day, and five men at $3 per day, 38,325.00 
 
 Total annual cost, $359,902.82 
 
 Annual cost of operating, and fixed charges per capita, $0.90 
 
 Or for filtration and sterilization of the drinking-water for a 
 city of 400,000 population, the cost per capita per annum cannot 
 be in excess of $1. Are we prepared to pay this for absolute 
 immunity from typhoid fever and other water-borne diseases ? 
 
 The Yaryan Company of New York has kindly furnished the 
 author an estimate of cost and operation for a water-sterilizing 
 plant upon its system, which will be more economical of fuel and 
 labor than the simple apparatus described. Adopting in our 
 estimate the figures supplied by this company, the costs are as 
 follows : - 
 
 COST OF CONSTRUCTION. 
 
 Filters and filter-house, $ 19,000.00 
 
 Yaryan quadruple effect sterilizer, 225,000.00 
 
 Boiler-house, 30,000.00 
 
 Pumping machinery and station, 33,000.00 
 
 Distributing mains, 1,575,000.00 
 
 Total, $1,882,000.00 
 
 FIXED ANNUAL CHARGES. 
 
 Interest and sinking-fund, $113,898.64 
 
 OPERATING EXPENSES. 
 
 100 tons of coal per day, at $2, for one year, $73,000.00 
 
 15 men at $2 per day, and 4 men at $3 per day, 15,330.00 
 
 Total annual cost, $202,228.64 
 
 Annual cost of operating and fixed charges per capita, $0.50 
 Or $.0554 per 1,000 gallons, sterilized and delivered. 
 
 This scheme for water purification involves, as shown, a sepa- 
 rate system of small mains to convey the sterilized and filtered 
 water from the works to the consumers, and requires a separate 
 
126 THE PURIFICATION OF WATER. 
 
 service pipe from these mains to bring the water into each prem- 
 ises, after which, as a measure of hygiene, the use of such water 
 for drinking and cooking purposes should, if found necessary, 
 be made compulsory. 
 
 It will be noticed that no allowance has been made in the cost 
 of operating for the cooling water to the surface condensers, because 
 the 97-98 per cent (or as much of it as may be required) of un- 
 sterilized water supplied to the city may be made to pass through 
 the condensers as cooling water without extra cost. 
 
 In regard to the figures heretofore given, it is not the purpose 
 to state with precision all the details of cost of this method of 
 water purification, but rather to lay down a principle, and let it be 
 worked out for each particular case. Doubtless in some cities the 
 cost of construction and operation will be less than has been shown, 
 while in others, for local reasons, it may be greater. But it is rea- 
 sonable to claim that sterilized and filtered water can be obtained 
 in our larger cities within the cost given, or at about one-tenth 
 cent per gallon ; and the purpose of the approximate figures stated 
 on the previous page is to show that the cost per unit of volume, 
 or per capita of population, for absolutely sterile water, is not so 
 great as to prohibit its use if demanded by the people. 
 
 Water such as this process of purification will furnish can 
 neither be the habitat nor carrier of any kind of bacteria (nor of 
 the toxalbumins which these may develop) ; and if any organisms 
 came into it adventitiously, they would perish for lack of food. 
 
 One great difficulty in the way of introducing a process for 
 the sterilization of drinking-water on a large scale in the United 
 States, lies in the well-known fact that the construction of public 
 works of any magnitude in most cities is seized upon as a political 
 advantage by the dominant party ; and the average tax-payer, upon 
 whom the burden of cost falls, usually views with alarm any prop- 
 osition to inaugurate an improvement requiring a large outlay of 
 money, in spite of the fact, perhaps, that his health or that of his 
 family, and possibly their lives, may depend upon the construction 
 of such works. In cities, however, like Berlin, Vienna, and St. 
 Petersburg, which are under imperial control, no difficulty should 
 be experienced in establishing a process for the sterilization and 
 
STERILIZATION OF DRINKING-WATER. 
 
 127 
 
 Side Elevation. 
 Fig. 4-. 
 
 Section Showing Circulation. 
 Fig. 5. 
 
 Figs. 3, 4, and 5. Yaryan Apparatus for Sterilizing Water. Quadruple Effect 
 
128 THE PURIFICATION OF WATER. 
 
 special distribution of the small percentage of water used for 
 drinking and culinary purposes whenever the health boards are 
 ready to recommend it. When this shall be done by these or any 
 other cities, and the sterilized water is used by all the citizens, 
 then such cities will be absolutely free from typhoid fever and 
 other water-borne diseases so far as these may be chargeable to 
 the local water supply. But no amount of care upon the part of 
 any city to defend its water supply from pollution, or render it safe 
 to health as a drinking-water after pollution, can prevent the im- 
 portation of typhoid fever from some other locality, where the 
 hygienic regulations for the drinking-water are less rigid, and 
 where the water contains the typhoid germ. From which we 
 reason that a system or process for the purification of drinking- 
 water, to be wholly effective, must be universal in its application ; 
 but no sanitary improvement, however essential it may be to 
 health, has been, or ever will be, applied everywhere at one and 
 the same time. It must have its origin in some city, the effica- 
 cious results must be shown and published ; whereupon other 
 cities, towns, and localities will speedily adopt it, and in due time 
 the benefits of such improvement will be enjoyed by all the civil- 
 ized people of the earth. 
 
 During the World's Fair at Chicago, 1893, all the employees, 
 numbering nearly 15,000, used sterilized drinking-water, with the 
 result that so long as this water only was used, no diarrheal 
 troubles were reported among the men. Upon the few occasions 
 when for short intervals of time the sterilized drinking-water was 
 discontinued, intestinal disorders arose ; and where typhoid fever 
 occurred, it was traced to a disregard of the rules of the Exposi- 
 tion with reference to drinking-water.* 
 
 The sterilization of the water was effected by passing it 
 through an ordinary feed-water heater, where it was raised to 212 
 Fahr., and kept at this temperature for a short time. Analysis of 
 the water revealed no bacterial life. 
 
 According to Surgeon-General Tryon of the United States 
 Navy,f " It may be stated that the medical officers of the navy 
 
 * Proceedings Fourteenth Annual Meeting A. W. W. Association, pp. 22-24. 
 
 t Water Supply, Chemical and Sanitary, Wm. P. Mason, New York, 1896, p. 156. 
 
STERILIZATION OF DRINKING-WATER. 129 
 
 recognize the great value of distilled water in the improvement in 
 health that has followed its introduction, particularly in certain 
 foreign stations." 
 
 No one will venture to deny that water properly sterilized by 
 heat in an ordinary steam-boiler will be absolutely safe for drink- 
 ing-purposes. All the bacteria or organic matter (in solution) 
 originally in such water will be wholly destroyed or precipitated 
 by evaporation under atmospheric pressure in a closed generator ; 
 and if we accept the proof that water is the carrier or original 
 cause of typhoid fever, then we are compelled to admit that water 
 properly sterilized cannot foster or carry the bacillus. With such 
 water universally used for drinking and cooking purposes, the 
 typhoid bacillus would perish, and typhoid fever cease to -exist. 
 
 With few exceptions, it is vain to look for water wholly safe 
 for drinking-purposes at its source. Few cities enjoy such water 
 to-day, otherwise among their inhabitants who use the water ex- 
 clusively typhoid fever would be unknown ; and doubtless in any 
 city where the water is of such quality that analysts have pro- 
 nounced it safe for drinking-purposes, it is being drunk to the 
 exclusion of any other available water. 
 
 The principal objections which have been offered to a double 
 water supply, whether the water of better quality is improved by 
 sterilization or filtration, or is naturally of high quality, are : 
 
 1. The better water will not always be used for drinking and 
 
 dietetic purposes. Some people will forget the danger 
 of drinking the unpurified water, and resort to that 
 which is most convenient, thereby defeating the very 
 purpose of a double supply. 
 
 2. It is urged that many of the poorer people cannot afford 
 
 to introduce two kinds of public water into their houses. 
 
 3. In many instances the better water will be used for other 
 
 than dietetic purposes. 
 
 The first of these objections can be overcome by proper edu- 
 cation and example. The instinct of self-preservation must be 
 wrought upon, and the natural tendency to take the better of two 
 things equally attainable may be expected to encourage the use of 
 
130 THE PURIFICATION OF WATER. 
 
 the better water for drinking and culinary purposes. The second 
 objection can be overcome by gentle .compulsion, in the same 
 manner that other sanitary improvements are carried out by mu- 
 nicipal corporations at the cost of the property benefited. The 
 third objection can readily be overcome by metering the purer 
 water supply, and in cases where such water is freely used about 
 the premises the cost thereof will be paid by the consumer. 
 
 It is probable that now a system of double water supply is 
 rather too refined for most municipal corporations, but it is possible 
 that such may be demanded by future generations. Many physi- 
 cians and hygienists at the present time favor the dual system, 
 through one branch of which water of the highest quality is to be 
 delivered for drinking and culinary purposes, while through the 
 other, water for the coarser uses is supplied to the consumers. 
 
 NOTE. Referring to the Yaryan Apparatus for the sterilization of drinking-water (p. 125), 
 quite extensive plants on this system have been in operation for several years at Perim and 
 Kosseir, on the Red Sea, and at Troon, Scotland ; converting from 6,000 to 12,000 gallons of 
 sea-water into fresh water per day of 24 hours, with an expenditure of about \ pound of coal 
 per gallon of water distilled, including the water for the boilers, and the fuel for operating the 
 necessary pumping-machinery. 
 
FILTRATION OF WATER SUPPLIES. 131 
 
 CHAPTER X. 
 
 FILTRATION OP WATER SUPPLIES. 
 
 THE purpose of this chapter is the discussion of devices and 
 plans for the purification of large volufnes of water for city use, 
 and is not intended to touch upon the subject of domestic filters. 
 The author believes that domestic filters, however well designed, 
 are, in the hands of the users, a delusion and a snare ; and instead 
 of being a safeguard against water-borne diseases, they really en- 
 courage the growth of the water bacteria, among which at times 
 may be pathogenic organisms. 
 
 The tests noted in Chapter III., on Bacterial Contents of 
 Various Waters, is therefore the only reference to domestic filters ; 
 and these have been included among others of water in and about 
 the city of Cincinnati. 
 
 Continuous sand filtration as practiced in Europe has gone 
 through an experience of nearly fifty years, and one would suppose 
 that this length of time should be sufficient to remove the matter 
 from the domain of experiment and establish it in the domain of 
 fact. Still, curiously enough, there are some who discuss sand fil- 
 tration as practiced abroad very much as they do the subject of 
 air navigation and the mobile perpetuum, things very interesting 
 in themselves, but quite impossible of any practical results. This 
 indifference to the wonderful performance of sand filters in Euro- 
 pean cities is a bar to the development of works of water purifica- 
 tion in this country, and is the cause of a large continuous loss of 
 valuable lives and much physical suffering, eighty to ninety per 
 cent of which might be averted if artificial works of water purifi- 
 cation were as largely used in this country as they are abroad. 
 
 Some writers in their enthusiasm have declared that sand filters 
 properly constructed and operated will furnish pure water. This 
 is a mistake. No filter operated upon a practical basis has ever 
 
132 THE PURIFICATION OF WATER. 
 
 furnished pure water ; but the so-called purified water is so much 
 superior to the unfiltered water that it will meet the practical 
 requirements of cities and communities to-day, and when the time 
 is reached that people demand absolutely pure water, methods for 
 furnishing it will doubtless be forthcoming. For the present, and 
 as a practical method of water purification, filtration may be re- 
 garded as entitled to full credit at the hands of city officials and 
 water-works managers. 
 
 Filtration, as the term is defined and generally understood, 
 consists of an interception or straining out from a fluid such sus- 
 pended matter as is larger in some dimension than the pores of 
 the filtering medium. The action is supposed to be purely mechan- 
 ical, and the efficiency of a filter will be measured by the fineness 
 or coarseness of the filtering material. The filtration of water, 
 however, demonstrates that the fineness of the filtering material 
 (sand) is not exactly a measure of the efficiency, and the finest 
 or smallest grain of sand does not always give the best results. 
 
 This fact, then, would naturally suggest that the straining 
 action is only a part of the work accomplished by the filter; and in 
 addition to the interception of certain suspended matters at the 
 surface of the sand-bed, some other forces are at work to reduce 
 the suspended matter, including the bacteria, in the water. One 
 of these forces is now known to be the action of the bacteria 
 on the organic matter. This is called the biologic action of the 
 filter. 
 
 All the common species of bacteria found in water are sapro- 
 phytes, and depend for subsistence on dead organic matter. In 
 fact, the bacteria are chiefly concerned in the destruction of this 
 organic matter, and its conversion into harmless nitrates. The 
 action of certain well-known forms of water bacteria upon sloped 
 agar, is seen to be the production of a film, or expansion so-called, 
 of its products of vital activity over the surface, and, if possessed 
 of anaerobic properties, in the body of the agar. This expansion 
 is indicative of the effect and propagation of the bacteria on the 
 food material. 
 
 The bulk of the suspended matter, including the bacteria in 
 water, will be intercepted at the surface of the sand. Here the 
 
FILTRATION OF WATER SUPPLIES. 133 
 
 process of splitting up the organic matter inio its nitrogenous 
 and carbonaceous elements is continually going on ; the carbons 
 going off as carbon dioxides and other gases, and the nitrogenous 
 matters being converted into nitrous and nitric acids, which in turn 
 unite with the bases in the water, forming nitrites and nitrates, in 
 themselves harmless products of bacterial action. 
 
 This biologic action of a filter is, after all, its most important 
 function. The simple straining process of a bed of sand or of 
 other filtering material, while competent to render turbid water 
 clear, could have but little effect upon the bacteria, because many 
 of these are so small in some dimension as to grow through a 
 sand-bed of almost any practicable fineness. The action of the 
 organisms in the water on the organic matter results in the pro- 
 duction of a thin semi-gelatinous film over and around the grains 
 of sand in the upper layers of a bed, which in due time becomes 
 so dense as to clog it, and require a high head to force the desired 
 amount of water through the sand ; whereupon such sand-bed is 
 temporarily taken out of service, the water drawn down some dis- 
 tance below the surface, and the upper fraction of an inch of the 
 sand removed. With the new surface of sand exposed, the filter 
 is ready for service again. 
 
 When the water is drawn off a filter for renewal of the surface 
 of the sand-bed, two important events occur. One consists of the 
 paring off of a thin layer of the clogged sand mentioned above ; 
 and the other of a complete or partial aeration of the sand-bed, by 
 means of which the nitrifying bacteria in the bed are supplied with 
 air (oxygen), without which, according to the authorities who have 
 especially studied these organisms,* the nitrifying bacteria would 
 soon perish, and their functions in the reduction of nitrogenous 
 organic matter to nitrous and nitric acids be lost. 
 
 All sand filters are therefore intermittent filters. None work 
 continuously. Each time the water is drawn down below the sur- 
 face of the sand-bed, there is a partial aeration of the sand; and 
 when the water is drawn off entirely, during the operation of par- 
 ing away the upper one-half () inch or so of dirty sand, the bed 
 is rested, as it were, and complete aeration occurs. 
 
 * Winogradsky, Warington, Percy Frankland, Dr. E. O. Jordan, and Mrs. Ellen H. Richards. 
 
134 THE PURIFICATION OF WATER. 
 
 This upper dirty layer of sand, which contains inorganic mat- 
 ter intercepted from the water, and the products of vital activity 
 of the water bacteria, is called the " Schmutzdecke " by Mr. Piefke, 
 who, the author believes, was the first to point out the manner in 
 which the semi-gelatinous film was formed, and how it consisted 
 of intercepted matter in suspension, and organic matter in process 
 of destruction by bacterial agency. 
 
 To one untutored in bacteriologic work, it may be difficult to 
 understand the action of the bacteria on nutrient matter in water ; 
 but to bacteriological students it is sufficient to state that in a fil- 
 ter the action of the bacteria upon suitable food material found in 
 the water will be like that of bacteria cultivated in sterilized arti- 
 ficial media. The materials found in water may be more or less 
 suitable for some of the water bacteria ; and those which find the 
 organic matter fitted to their needs will flourish on the surface of 
 a sand-bed, and appropriate to their support such matter as may be 
 found in suspension or intercepted at or near the surface of the 
 sand. 
 
 It is abundantly proven that the bacteria do not penetrate the 
 sand-bed to any great depth,* and the surface of the sand where 
 the interception of suspended matter must occur is also the prin- 
 cipal seat of operations of the bacteria and other organisms in the 
 water. The bacteria are not the only forms of life in water, and 
 some allowance must be made for the destructive action of the 
 infusoria and other forms of aquatic life upon the organic matter 
 of which all these forms are themselves a part. 
 
 The following diagram and description showing the rate at 
 which the bacteria grow in the sand-bed from the surface down- 
 ward, is taken from Mr. Gill's paper on the filters at the Freder- 
 ickshagen Station of the Berlin Water-Works. 
 
 "It has been stated above that the number of bacteria colonies is greatest 
 at the surface of the sand, and decreases very rapidly in successive layers be- 
 neath. In Fig. G, 0, .r, z, y, represents the 2-foot deep sand-layer of a filter. 
 If with as origin, and distances along 0, x, representing depths of sand-layer, 
 and those along 0, j/, numbers of bacteria colonies per kilogram of sand, the 
 
 * The Filtration of the Miiggel Lake Water Supply, Berlin, by Henry Gill, M.I.C.E., 
 London, 1895, p. 12. 
 
FILTRATION OF WATER SUPPLIES. 
 
 135 
 
 60,000,000 bacteria colonies per kilogram of the ripe sarfd be plotted at each 
 of the depths 0, 4, 8, 12, and 24 inches, the line h, w, parallel to 0, x is arrived 
 at. The hatched strip 0, x, m, h, then represents the ripe condition of filter 
 sand after long use, in which condition a powerful water current and attrition 
 of the grains against each other fail to free them from the bacteria. If now, 
 with the same abscissae, the 734, 190, 150, 92, and 60 millions be plotted as 
 ordinates, the curve m, p, q, r, is arrived at. This curve exhibits pictorially 
 the density of the bacteria colonies in the various layers of a sand filter at the 
 close of a period of service when it gives the best results." 
 
 Surface of Sand Bed 
 
 24" Sand Bed 
 
 <5and 
 
 tnin 
 i 
 
 i 
 
 Gravel 
 Fig. 6. Diagram Showing Accumulation of Bacteria in Sand-Bed. 
 
 The "Schmutzdecke," or film of intercepted suspended matter 
 and products of bacterial action, is a delicate membrane lacking in 
 consistency, and easily broken by too rapid changes of pressure 
 (head) on the sand-bed ; and when broken bad results are liable to 
 follow. The author cannot do better than quote again from Mr. 
 Gill upon this feature of sand filtration.* 
 
 " Since the bacteria are liable to be washed downwards by a stream of 
 greater force than that which prevailed when they came into contact with the 
 sand grains, it is of the utmost importance to avoid an increase of speed, 
 especially a sudden increase. Mechanical arrangements must be adopted to 
 prevent this, and it must be impossible that any filter in action can in any way 
 affect the yield of the neighboring filter. The chief cleansing action takes 
 place in the mud deposit on the surface of the sand, and in the sand immediately 
 at the surface. In this region the coating of deposit is soft, and with its dense 
 population requires careful and tender treatment to avoid squeezing out the 
 bacteria by undue pressure. It is obvious that as soon as an appreciable 
 
 * The Filtration of the Miiggel Lake Water Supply, Berlin, by Henry Gill, M.I.C.E., 
 London, 1895, p. 12. 
 
136 THE PURIFICATION OF WATER. 
 
 deposit has taken place on the sand surface, any increase of ' head ' must be 
 chiefly caused by the layer of this deposit. If the sand beneath is not abso- 
 lutely homogeneous, as it cannot be, any increase of pressure may cause a 
 depression and a tearing of the mud-skin on the less dense parts of the surface 
 of the sand. Through such a rupture the bacteria are at once washed by the 
 increased local current which ensues into the sand beneath, and may be carried 
 through the entire layer. Yet a gradual increase of pressure must of necessity 
 take place, when the yield is to be constant, in order to overcome the increas- 
 ing friction of the passage of the water through the filtering medium, in pro- 
 portion as its insterstices become gradually closed by the deposit. Nor is such 
 increase, if gradual, injurious, provided certain limits be not exceeded. 
 
 Mr. Gill states that the maximum " head " for the Miiggel Lake 
 filter is 2 feet, but this has been increased in other instances to 
 5 or 6 feet without an apparent breaking of the surface film on 
 the sand-bed. 
 
 After a filter has been scraped, and refilled with filtered water 
 from below to the surface of the sand, water should r3e drawn from 
 the settling-basin onto the filter to the full depth of high-water 
 mark, and allowed to stand several hours before any flow occurs 
 from the filter. This interval of rest will be in continuation of the 
 subsidence of the suspended matter in the water, and will assist in 
 the formation of a coat of slime over the sand-bed before the flow 
 is started. After the water has remained at rest for a few hours 
 over the sand-bed, the flow should be started cautiously, at a low 
 rate, and gradually increased until the maximum allowable rate has 
 been attained. 
 
 ACTION OF THE INTERMITTENT SAND FILTER. 
 
 In the continuous sand filter substantially the whole work of 
 purification is accomplished at or near the surface of the sand-bed. 
 The " Schmutzdecke," or dirty cover, which is regarded by foreign 
 engineers as an essential of proper eand filtration, is not consid- 
 ered of special importance in the intermittent filter. In this it is 
 assumed that the organic matter in the water is reduced by the 
 action of the bacteria in the bed of sand to nitrous and nitric acids, 
 which unite with the bases in the water, forming insoluble and 
 harmless nitrites and nitrates; This work is chiefly accomplished 
 by the nitrifying bacteria, discovered by Winogradsky in the soil 
 
FILTRATION OF WATER SUPPLIES, 137 
 
 at Zurich, and by Dr. Jordan and Mrs. Richards in the sewage 
 at Lawrence, Mass. ; but the ordinary water bacteria are also useful 
 in breaking up the organic matter in water before it is acted upon 
 by the nitrifiers. 
 
 The intermittent filter receives the water for a number of hours 
 or days, and then rests for a number of hours or days, until the 
 water held in the sand-bed has drained away, and the interstitial 
 spaces are filled with air, when the filter is supplied with water as 
 before. 
 
 Thus, while the action is intermittent from day to day, it other- 
 wise is continuous in operation, the scraping and cleaning of the 
 sand being only such as ca readily be done during the short 
 intervals of rest. 
 
 In the intermittent sand filter the bed of sand for the whole 
 depth is supposed to act in the work of water purification, while 
 the upper one-half inch or less of sand is known in the continuous 
 filter to be concerned in the reduction of the bacterial contents as 
 well as of other matters held in suspension by the water. The 
 operation of the Lawrence, Mass., filter, since it was put in regular 
 service, suggests that there has been some departure from the 
 original method of use, and that in the days of continual service, 
 and the resting and scraping of the sand-bed, it conforms mose 
 nearly than was intended to the method of operation pursued with 
 sand filters abroad. 
 
 By the thorough and frequent aeration of the sand-bed, it is held 
 by the designer (Mr. Mills) that there will be a "burning up" of 
 the organic matter intercepted at the surface and in the depth 
 of the filter, and by proportioning properly the duration of ser- 
 vice and rest, with a complete draining of the bed each time it is 
 rested, all organic matter will be consumed. The aeration of the 
 sand-bed is intended to maintain the vitality of the nitrifying 
 bacteria, which are the organisms concerned in the final destruc- 
 tion of the organic matter, and its conversion into nitrites and 
 nitrates. 
 
 It is the theory that the intermittent filter is capable of con- 
 tinuous renewal by the forces within itself, and the large periodical 
 expense required to restore the sand-bed of the continuous filter to 
 
138 
 
 THE PURIFICATION OF WATER. 
 
 its normal condition will be materially reduced. Enough experi- 
 ence has not been had with this system of filtration to express an 
 opinion upon its adaptability to other waters than that of the Mer- 
 rimac River ; but both in the bacterial results and typhoid rates of 
 Lawrence, since it was put in service four years ago, it seems not 
 to have attained the high standard of efficiency reached by the 
 continuous sand filters of Europe. 
 
 Any statement heretofore made upon the operation of sand 
 filters is assumed to be the natural action without the aid of ex- 
 traneous materials to assist in the precipitation of suspended mat- 
 ters, or in the formation of a coagulum at the surface of the sand. 
 Aside from the use of particles of iron in a revolving cylinder 
 (Anderson process), it is not known that any artificial agency is 
 relied upon to insure the successful operation of sand filters in 
 Europe. 
 
 From a knowledge of natural filtration as it occurs in the drift, 
 it is easy to perceive that artificial sand filtration may be made to 
 accomplish results far superior to natural filtration as it some- 
 times occurs. Thus the size and uniformity of the sand-grains, the 
 effective head, and rate of flow through the sand-bed, may be so 
 proportioned that the resultant filtrate is equal in purity to spring 
 or deep well water. And this result can be obtained, not seldom, 
 but at all times, and without regard to the original condition of the 
 water. Assuming that the typhoid fever death rate is a correct 
 index of the quality of a public water supply, then it appears that 
 filtration can produce a water which will rival the purest of natural 
 waters. 
 
 The water of Vienna and Munich is mountain spring water, 
 not surpassed by any, and equaled by that of few cities of the 
 world. The typhoid rates by the author's scale (Chapter V.) for 
 these cities since 1890 have been: 
 
 DEATH RATES FROM TYPHOID FEVER PER 100,000 OF POPULATION LIVING. 
 
 YEARS, 
 
 1890. 
 
 1891. 
 
 1892. 
 
 1893. 
 
 1894. 
 
 1895. 
 
 1896. 
 
 AVERAGE. 
 
 Vienna, 
 
 9 
 
 6 
 
 8 
 
 7 
 
 5 
 
 6 
 
 5 
 
 6.55 
 
 Munich, 
 
 8 
 
 7 
 
 8 
 
 15 
 
 2.5 
 
 3 
 
 3 
 
 5.94 
 
FILTRATION OF WATER SUPPLIES. 
 
 139 
 
 Omitting the rate for 1893 for Munich, when there appears 
 to have been an unwarranted increase in the typhoid rates, the 
 average for the other six years becomes 4.4. 
 
 The water of Rotterdam and Berlin is filtered, the first from 
 the River Maas, and the second from the Rivers Spree and Havel, 
 both of which have received sewage and surface drainage from 
 urban and rural territory before the water reaches the intakes of 
 these works. 
 
 DEATH RATES FROM TYPHOID FEVER PER 100,000 OF POPULATION LIVING. 
 
 YEARS, 
 
 1890. 
 
 1891. 
 
 1892. 
 
 1893. 
 
 1894. 
 
 1895. 
 
 1896. 
 
 AVERAGE. 
 
 Rotterdam, 
 
 6 
 
 4 
 
 6 
 
 5 
 
 4.8 
 
 2 
 
 12 
 
 5.7 
 
 Berlin, 
 
 9 
 
 10 
 
 8 
 
 9 
 
 4 
 
 5 
 
 5 
 
 7.14 
 
 Note the fact that the death rate for Rotterdam is lower than 
 for either Vienna or Munich. Note also that the comparison is not 
 between cities of one country where the consumption of beer and 
 other beverages is high, and of another country where from mod- 
 esty, if for no other reason, we must claim that the consumption 
 of beverages other than water is low. 
 
 Aside from the fact of equal quality as shown by comparison 
 of spring and filtered waters, the theory of sand filtration, properly 
 studied, leads to the conclusion that water of more uniform quality 
 can be had from artificial filters than from irregular and scattered 
 sand-beds as found in the drift, in some of which the size and 
 irregularity of sand-grain and position of the sand-bed are not cal- 
 culated for proper filtration. Spring water and well water may be 
 pure, but we cannot state with assurance how it has been made 
 pure ; while with filtered water we know how purity, so-called, has 
 been obtained, and by repeating the process of purification we can 
 reproduce the quality of filtrate. 
 
 The insufficiency of natural filtration through the drift is well 
 recognized by those who have given the matter serious considera- 
 tion. Dr. Drown, in the Report of the Massachusetts State Board 
 of Health for 1891, p. 355, says: 
 
 " Although water badly contaminated with sewage or the wastes of human 
 life may be purified by thorough nitration so as to be free from organic matter 
 
140 THE PURIFICATION OF WATER. 
 
 and bacteria, yet in cases of ground waters of this origin and character we sel- 
 dom feel complete security that the conditions of perfect filtration will always 
 exist. A long-continued rainfall, for instance, may result in more rapid filtra- 
 tion, and consequently less perfect purification ; or the creation of new sources 
 of contamination nearer the spring may result in its dangerous pollution. 
 
 "It is for such reasons that a certain suspicion always attaches to ground 
 waters which have at any time in their history been seriously polluted. The 
 use of ground waters, whether springs or wells, in built-up communities, should 
 therefore be avoided ; for we have no control over the conditions of filtration, 
 and have no means of knowing (except by constant vigilance in the examina- 
 tion of the water) when a water hitherto well purified may become injuriously 
 impure. The danger from the use of ground waters in populous regions in- 
 creases v/ith the increase of population, and with the nearness of the sources 
 of pollution to the spring or well." 
 
 The methods of water purification which have given such 
 excellent results in cities of Europe are generally sedimentation 
 for a few days in large reservoirs, combined with slow filtration 
 through beds of sand ; and in some situations, like that of Berlin 
 at Lake Miiggel, where there is usually but little turbidity to the 
 water, it is at times pumped direct from the lake to the filters. 
 
 Sedimentation is accomplished in reservoirs which will hold 
 from a day to several days' supply. While at a state of rest in 
 these reservoirs, much of the suspended matter which imparts 
 color to the water will be precipitated, and form layers of mud on 
 the bottom and sides of the basin. 
 
 While a few days' subsidence of turbid polluted water may 
 have no large influence upon its quality, it will remove much of 
 the suspended matter which otherwise will clog a sand filter, and 
 reduce its term and efficiency of service. 
 
 Careful study of the subject of sand filtration has led to the 
 opinion that it is possible to have the sand so fine and the rate 
 of filtration so slow that, theoretically, all suspended matter, in- 
 cluding the bacteria will be arrested on or in the sand-bed. But 
 this would require enormous areas of filter surface, with limited 
 commercial efficiency, and the cost of water so obtained would 
 be prohibitory on a large scale. 
 
 In order, however, to approach as nearly as practicable the 
 ideal condition of filtrate at a reasonable cost, it is desirable that 
 
FILTRATION OF WATER SUPPLIES. 141 
 
 all the heavier matter should be removed by subsidence before the 
 water is put on the filters. Hence the use, in the water-works of 
 London, Hamburg, and other foreign cities, of subsiding reservoirs 
 in which the water is stored for several hours or days before the 
 process of filtration begins. 
 
 It is not possible to make a sewage-polluted water fit for 
 drinking-purposes by subsidence alone, excepting the water is 
 permitted to remain in a wholly quiescent state in large, deep 
 reservoirs for many months or years ; a condition altogether im- 
 practicable for most cities ; while subsidence for a few hours or 
 days will reduce the suspended matter and silt in most turbid 
 waters to a state which will admit of the use of comparatively fine 
 sand in the filters, and rates of delivery higher than the average 
 of European practice. The Engineer Commission on the Im- 
 provement of the Water-Works of the city of Cincinnati, proposed 
 a rest of the Ohio-river water in subsiding reservoirs for four days 
 before it was drawn off to the filters. 
 
 The time allowed for sedimentation before the water is thrown 
 on the filters varies in different cities, and sometimes is controlled 
 by financial rather than hygienic considerations. The following 
 table contains the data upon this subject, from a few of the works 
 abroad which combine subsidence with filtration : 
 
 TIME ALLOWED FOR SEDIMENTATION. 
 
 LONDON, 
 
 Chelsea Works, 
 
 12.0 days. 
 
 
 West Middlesex, 
 
 5.6 
 
 
 Southwark, 
 
 4.1 
 
 
 Grand Junction, 
 
 3.3 
 
 
 Lambeth, 
 
 6.0 
 
 
 New River, 
 
 4.4 
 
 
 East London, 
 
 15.0 
 
 HAMBURG, 
 
 
 19-30 hours. 
 
 ROTTERDAM, 
 
 
 24 
 
 BERLIN, Frederickshagen Works, 24 
 
 The views of English engineers at present distinctly favor sed- 
 imentation of surface waters previous to filtration ; and the new 
 works proposed by Sir A. R. Binnie for the supply of London, 
 notwithstanding the water impounded from the Welsh sources is 
 
142 THE PURIFICATION OF WATER. 
 
 naturally of very high quality, contemplates nitration of this water 
 before it is distributed to the consumers. 
 
 In his report on the new sources of supply proposed for 
 London, Mr. Binnie says : 
 
 " Although it will be seen from the chemist's analyses of the water of the 
 Usk, the Yrfon, the Towy, the Wye, etc., that the waters in their natural state 
 are of greater purity and contain less solid matter than the London water 
 after nitration, and although these waters will be stored and be subjected to 
 subsidence in the large reservoirs which I have described, and in some cases 
 will be decanted or drawn off from one reservoir into another, yet I consider 
 that when all precautions are taken, the water should be filtered before deliv- 
 ering to the consumer." 
 
 According to certain principles formulated by the Imperial 
 Board of Health, Berlin (1893), the rate of nitration should not 
 exceed 4 inches vertical .per hour, or 8 feet per day, which corre- 
 sponds to a daily rate per acre of 2,606,630 U. S. gallons. From 
 some experiments by the late Mr. W. Kiimmel, engineer of the 
 Altona, Germany, Water- Works,* at rates of filtration of 4, 8, and 
 16 feet vertical per day, he obtained the best bacterial results from 
 the higher rates, as indicated by the following table : 
 
 1,303,315 U. S. gallons per acre per day = 11 to 97 colonies per c. c. 
 2,606,630 = 5 to 79 
 
 5,213,260 " = 7 to 72 
 
 Mr. Kiimmel did not regard 8 feet per day as "beyond doubt 
 the maximum of safe filtration;" he thought though, that "the 
 danger of a trespassing pathogenic organism is much more unlikely 
 at the lower than at the higher rates, and that the best velocity 
 was not the same for all waters." He felt confident "that the 
 difference in the mineral, vegetable, and animal admixtures is of 
 high importance in this question," and that we should endeavor to 
 ascertain the best rate for each separate water and water-works. 
 
 From the latest published Annual Report of the Massachusetts 
 State Board of Health (1895), the following notes from the exper- 
 imental filters at the Lawrence station are taken : 
 
 * Transactions American Society of Civil Engineers, vol. xxx., p. 333. 
 
FILTRATION OF WATER SUPPLIES. 
 
 143 
 
 INTERMITTENT SAND FILTERS. 
 
 
 
 AVERAGE 
 
 
 RATE OF FILTRATION. 
 
 BACTERIA 
 
 BACTERIA 
 
 PERCENTAGE OF 
 
 GALLONS PER ACRE PER DAY. 
 
 PER C. C. IN 
 
 PER C. C. IN 
 
 BACTERIA 
 
 
 RIVER WATER. 
 
 FILTRATE. 
 
 REMOVED. 
 
 2,000,000 
 
 11,600 
 
 29 
 
 99.75 
 
 2,500,000 to 5,000,000 
 
 16,300 
 
 137 
 
 99.16 
 
 5,000,000 to 7,000,000 
 
 11,600 
 
 72 
 
 99.38 
 
 CONTINUOUS SAND FILTERS. 
 
 
 AVERAGE 
 
 AVERAGE 
 
 
 RATE OF FILTRATION. 
 
 BACTERIA 
 
 BACTERIA 
 
 PERCENTAGE OF 
 
 GALLONS PER ACRE PER DAY. 
 
 PER C. C. IN 
 
 PER C. C. IN 
 
 BACTERIA 
 
 
 RIVER WATER. 
 
 FILTRATE. 
 
 REMOVED. 
 
 1,000,000 to 2,500,000 
 
 13,950 
 
 72 
 
 99.49 
 
 2,500,000 to 5,000,000 
 
 18,220 
 
 273 
 
 98.56 
 
 5,000,000 to 7,000,000 
 
 11,600 
 
 73 
 
 99.37 
 
 7,000,000 to 10,000,000 
 
 16,500 
 
 130 
 
 99.22 
 
 The average results given in the table were obtained with sands 
 varying in " effective size" from 0.14 to 0.48 mm., and "uniform- 
 ity coefficient " from 1.6 to 3.7, while the thickness of sand-bed 
 varied from 60 to 7 inches. Considering the efficiencies of the 
 filters with sand-beds not less than 48 inches in thickness, the 5 
 intermittent filters for rates of filtration from 2,000,000 to 6,600,- 
 000 gallons per acre per day gave an average bacterial reduction 
 of 99.32 per cent, while the 8 continuous filters for rates of filtra- 
 tion from 2,000,000 gallons to 8,200,000 gallons per acre per day 
 gave an average bacterial reduction of 99.32 per cent.* 
 
 The influence of rate of filtration on the organic matter and 
 bacterial contents of the Zurich filtered water is shown by Dr. A. 
 Bertschinger of the Municipal Laboratory of Zurich, in the follow- 
 ing table : 
 
 CHEMICAL QUANTITIES 
 
 RATE OF FILTRATION PER ACRE PER DAY IN U. S. GALLONS. 
 
 FILTERED WATER. 
 
 4,356,000. 
 
 5,227,200. 
 
 8,712,000. 
 
 16,262,400. 
 
 21,489,600. 
 
 Organic matter, 
 
 1.65 
 
 1.69 
 
 1.70 
 
 1.70 
 
 2.02 
 
 Free ammonia, 
 
 0.0007 
 
 0.0008 
 
 0.0004 
 
 0.0004 
 
 0.0006 
 
 Albuminoid ammonia, 
 
 0.0028 
 
 0.0027 
 
 0.0027 
 
 0.0027 
 
 0.0037 
 
 Bacteria per c. c. in filtrate, 
 
 20 
 
 31 
 
 22 
 
 15 
 
 18 
 
 * Twenty- seventh Annual Report Massachusetts State Board of Health, p. 505. 
 
144 
 
 THE PURIFICATION OF WATER. 
 
 Commenting on these results, Mr. Preller * says (adapting his 
 figures to U. S. gallons and rates per acre per day) : 
 
 " These results show, therefore, that, provided the filter-beds are in efficient 
 working order, neither the chemical nor the bacteriological purity of the filtered 
 water is impaired by increasing the rate of percolation from 5,953,200 to 
 16,262,400 U. S. gallons per acre per day, a fact which is at variance with the 
 view advanced elsewhere, that the mean rate of percolation for sand-filters 
 should be limited to 3,194,400 U. S. gallons per acre per day." 
 
 According to Mr. Schroder, the number of bacteria in the un- 
 filtered Elbe water at Hamburg ranges from 800 to 3,000, while 
 the filtered water seldom contains above 30 colonies per cubic 
 centimeter, and at times is as low as 20 colonies per cubic cen- 
 timeter, showing a reduction of 97.5 to 99.0 per cent in the 
 bacterial contents of the raw water. 
 
 The following tables from Dr. E. Frankland's f bacterial an- 
 alysis of the water supplied by the London companies from the 
 Rivers Thames and Lea, are very interesting when viewed from 
 the standpoint of artificial water purification upon a large scale. 
 
 COMPANIES WHICH TAKE WATER FROM THE RIVER THAMES. 
 CHELSEA WATER COMPANY. 
 BACTERIA PER C. C. OF WATER. 
 
 MONTH. 
 
 UNFILTERED 
 WATER. 
 
 AFTER 12 DAYS' 
 STORAGE. 
 
 AFTER 
 FILTRATION. 
 
 January, 
 
 11,560 
 
 1,360 
 
 20 
 
 February, 
 
 26,800 
 
 460 
 
 44 
 
 March, 
 
 18,000 
 
 240 
 
 28 
 
 April, 
 
 7,520 
 
 Lost. 
 
 4 
 
 May, 
 
 2,060 
 
 140 
 
 24 
 
 June, 
 
 6,760 
 
 1,150 
 
 178 
 
 July, 
 
 2,220 
 
 420 
 
 20 
 
 August, 
 
 1,740 
 
 200 
 
 18 
 
 September, 
 
 4,300 
 
 140 
 
 2 
 
 October, 
 
 39,760 
 
 340 
 
 8 
 
 November, 
 
 8,560 
 
 280 
 
 12 
 
 December, 
 
 160,000 
 
 854 
 
 55 
 
 Average, 
 
 24,107 
 
 508 
 
 34 
 
 Average percentage of reduction by subsidence, 97.85 
 
 Average percentage of reduction by subsidence and filtration, 99.86 
 
 * Zurich Water Works, C. P. Du R. Preller, London, 1892, p. 26. 
 
 t Annual Summary of Vital Statistics, London, 1896, p. Ixxiv., et seq. 
 
FILTRATION OF WATER SUPPLIES. 
 
 145 
 
 WEST MIDDLESEX COMPANY. 
 BACTERIA PER C. C. OF WATER. 
 
 MONTH. 
 
 UNFILTERED 
 WATER. 
 
 AFTER 5.6 DAYS' 
 STORAGE. 
 
 AFTER 
 FILTRATION. 
 
 January, 
 
 11,560 
 
 3,460 
 
 44 
 
 February, 
 
 26,800 
 
 1,820 
 
 16 
 
 March, 
 
 18,000 
 
 2,340 
 
 24 
 
 April, 
 
 7,520 
 
 720 
 
 20 
 
 May, 
 
 2,060 
 
 280 
 
 4 
 
 June, 
 
 . 6,760 
 
 1,000 
 
 301 
 
 July, 
 
 2,220 
 
 680 
 
 8 
 
 August, 
 
 1,740 
 
 300 
 
 6 
 
 September, 
 
 4,300 
 
 120 
 
 14 
 
 October, 
 
 39,760 
 
 740 
 
 30 
 
 November, 
 
 8,560 
 
 5,520 
 
 25 
 
 December, 
 
 160,000 
 
 26,760 
 
 120 
 
 Average, 
 
 24,107 
 
 3,605 
 
 si 
 
 Average percentage of reduction by subsidence, 85.05 
 
 Average. percentage of reduction by subsidence and nitration, 99.79 
 
 SOUTHWARK AND VAUXHALL COMPANY. 
 BACTERIA PER C. C. OF WATER. 
 
 MONTH. 
 
 UNFILTERED 
 WATER. 
 
 AFTER 4.1 DAYS' 
 STORAC.E. 
 
 AFTER 
 FILTRATION. 
 
 January, 
 
 11,560 
 
 . . 
 
 32 
 
 February, 
 
 26,800 
 
 
 234 
 
 March, 
 
 18,000 
 
 . . 
 
 102 
 
 April, 
 
 7,520 
 
 . . 
 
 1,116 
 
 May, 
 
 2,060 
 
 
 36 
 
 June, 
 
 6,760 
 
 . . 
 
 24 
 
 July, 
 
 2,220 
 
 . . 
 
 188 
 
 August, 
 
 1,740 
 
 . . 
 
 12 
 
 September, 
 
 4,300 
 
 . . 
 
 68 
 
 October, 
 
 39,760 
 
 . . 
 
 16 
 
 November, 
 
 8,560 
 
 
 142 
 
 December, 
 
 160,000 
 
 920 
 
 8,020 
 
 Average, 
 
 24,107 
 
 
 832.5 
 
 Average percentage of reduction by filtration alone, 
 
 96.55 
 
146 
 
 THE PURIFICAl^ION OF WATER. 
 
 GRAND JUNCTION COMPANY. 
 BACTERIA PER C. C. OF WATER. 
 
 MONTH. 
 
 UNFILTERED 
 WATER. 
 
 AFTER 3.3 DAYS' 
 STORAGE. 
 
 AFTER 
 FILTRATION'. 
 
 January, 
 
 11,560 
 
 290 
 
 28 
 
 February, 
 
 26,800 
 
 400 
 
 83 
 
 March, 
 
 18,000 
 
 380 
 
 97 
 
 April, 
 
 7,520 
 
 1,110 
 
 112 
 
 May, 
 
 2,060 
 
 540 
 
 56 
 
 June, 
 
 6,760 
 
 567 
 
 376 
 
 My, 
 
 2,220 
 
 410 
 
 32 
 
 August, 
 
 1,740 
 
 510 
 
 21 
 
 September, 
 
 4,300 
 
 360 
 
 63 
 
 October, 
 
 39,760 
 
 740 
 
 49 
 
 November, 
 
 8,560 
 
 1,580 
 
 110 
 
 December, 
 
 160.000 
 
 38,000 
 
 1,106 
 
 Average, 
 
 24,107 
 
 3,741 
 
 178 
 
 Average percentage of reduction by subsidence, 84.48 
 
 Average percentage of reduction by subsidence and filtration, 99.20 
 
 LAMBETH COMPANY. 
 BACTERIA PER C. C. OF WATER. 
 
 MONTH. 
 
 UNFILTERED 
 WATER. 
 
 AFTER 6.0 DAYS' 
 STORAGE. 
 
 AFTER 
 FILTRATION. 
 
 January, 
 
 11,560 
 
 6,560 
 
 56 
 
 February, 
 
 26,800 
 
 13,380 
 
 56 
 
 March, 
 
 18,000 
 
 5,120 
 
 40 
 
 April, 
 
 7,520 
 
 5,340 
 
 12 
 
 May, 
 
 2,060 
 
 1,080 
 
 8 
 
 June, 
 
 6,760 
 
 1,280 
 
 130 
 
 July, 
 
 2,220 
 
 1,340 
 
 20 
 
 August, 
 
 1,740 
 
 600 
 
 60 
 
 September, 
 
 4,300 
 
 1,080 
 
 30 
 
 October, 
 
 39,760 
 
 4,660 
 
 12 
 
 November, 
 
 8,560 
 
 2,920 
 
 24 
 
 December, 
 
 160,000 
 
 56,000 
 
 116 
 
 Average, 
 
 24,107 
 
 8,280 
 
 47 
 
 Average percentage of reduction by subsidence, 65.65 
 
 Average percentage of reduction by subsidence and filtration, 99.81 
 
FILTRATION OF WATER SUPPLIES. 
 
 147 
 
 COMPANIES WHICH TAKE WATER FROM THE fclVER LEA. 
 
 NEW RIVER COMPANY. 
 BACTERIA PER C. C. OF WATER. 
 
 MONTH. 
 
 UNFILTEKED 
 WATER. 
 
 AFTER 4.4 DAYS" 
 STORAGE. 
 
 AFTER 
 FILTRATION. 
 
 January, 
 
 2,510 
 
 1,040 
 
 31 
 
 February, 
 
 2,080 
 
 1,580 
 
 31 
 
 March, 
 
 4,240 
 
 1,820 
 
 11 
 
 April, 
 
 1,340 
 
 500 
 
 4 
 
 May, 
 
 1,340 
 
 300 
 
 7 
 
 June, 
 
 1,640 
 
 420 
 
 17 
 
 Jiy, 
 
 1,500 
 
 480 
 
 12 
 
 August, 
 
 840 
 
 340 
 
 67 
 
 September, 
 
 2,540 
 
 600 
 
 16 
 
 October, 
 
 4,400 
 
 820 
 
 7 
 
 November, 
 
 3,200 
 
 4,880 
 
 69 
 
 December, 
 
 14,540 
 
 7,480 
 
 266 
 
 Average, 
 
 3,347 
 
 1,693 
 
 45 
 
 Average percentage of reduction by subsidence, 49.42 
 
 Average percentage of reduction by subsidence and filtration, 98.65 
 
 EAST LONDON COMPANY. 
 BACTERIA PER C. C. OF WATER. 
 
 MONTH. 
 
 UNFILTERED 
 WATER. 
 
 AFTER 15 DAYS' 
 STORAGE. 
 
 AFTER 
 FILTRATION. 
 
 January, 
 
 6,720 
 
 3,140 
 
 68 
 
 February, 
 
 7,880 
 
 1,600 
 
 69 
 
 March, 
 
 20,640 
 
 1,460 
 
 49 
 
 April, 
 
 Lost. 
 
 Lost. 
 
 52 
 
 May, 
 
 8,180 
 
 1,180 
 
 81 
 
 June, 
 
 11,720 
 
 2,340 
 
 208 
 
 July, 
 
 2,680 
 
 1,520 
 
 43 
 
 August, 
 
 6,020 
 
 2,140 
 
 68 
 
 September, 
 
 32,000 
 
 2,160 
 
 41 
 
 October, 
 
 12,220 
 
 1,460 
 
 53 
 
 November, 
 
 10,880 
 
 3,200 
 
 62 
 
 December, 
 
 80,000 
 
 13,420 
 
 145 
 
 Average, 
 
 18,085 
 
 3,056 
 
 78 
 
 Average percentage of reduction by subsidence, 83.10 
 
 Average percentage of reduction by subsidence and filtration, 99.56 
 
148 
 
 THE PURIFICATION OF WATER. 
 
 PERCENTAGE OF BACTERIA REMOVED. 
 
 WATER COMPANY. 
 
 No. OF DAYS OF 
 SUBSIDENCE. 
 
 BY 
 
 SUBSIDENCE. 
 
 BY SUBSIDENCE 
 AND FILTRATION. 
 
 Grand Junction, 
 
 3.3 
 
 84.48 
 
 99.26 
 
 New River, 
 
 4.4 
 
 49.42 
 
 98.65 
 
 West Middlesex, 
 
 5.6 
 
 85.05 
 
 99.79 
 
 Lambeth, 
 
 6.0 
 
 65.65 
 
 99.81 
 
 Chelsea, 
 
 12,0 
 
 97.85 
 
 99.86 
 
 East London, 
 
 15.0 
 
 83.10 
 
 99.56 
 
 Southwark and 
 
 Percentage of Bacteria removed by Fil- 
 
 
 Vauxhall, 
 
 . tration without Subsidence. 
 
 96.55 
 
 The preceding tables have been given in some detail in order 
 to discuss the numbers of bacteria in the water after filtration. 
 The first observation which one will naturally make is the extreme 
 variation of results for different months by the same company, 
 and for the same months by the different companies. Keeping in 
 view the London standard of bacterial contents of potable water, 
 i.e., 100 colonies per cubic centimeter, it appears that all of 
 the companies complied with the standard for the month of Janu- 
 ary. Only one of the seven companies (Southwark) failed to 
 comply with the required standard for February ; and only one 
 company, the Southwark again, failed to bring the bacterial con- 
 tents of the filtrate within the prescribed limit for the month of 
 March. 
 
 The record of the Southwark Company indicates very bad 
 work for several months, and as an average for the year, and is 
 to be accounted for only upon the ground of insufficient filter 
 capacity, or gross negligence in the manipulation of the filters. 
 Once only did the Chelsea and New River Companies, and twice 
 only during the year did the West Middlesex, Lambeth, and East 
 London Companies pass the bacterial limit, while the work of the 
 Southwark and Grand Junction Companies for the year was gen- 
 erally very poor. 
 
 With the exception of the Chelsea Company, the work of the 
 filters for December was not up to the standard of London water. 
 Referring to the Chelsea Company, and omitting the bad work of 
 the filters for June, the average for the other eleven months was 
 
FILTRATION OF WATER SUPPLIES. 149 
 
 21.4 bacteria per cubic centimeter, of filtered water, at times fall- 
 ing so low as 2 and never exceeding 55. Neglecting the bad work 
 of the West Middlesex filters for the months of June and Decem- 
 ber, the average for the other ten months was 19.1, the lowest 
 count being 4, and the highest count 44 bacteria per cubic centi- 
 meter. 
 
 An examination of all the tables reveals the fact that four of 
 the companies at times brought the bacterial condition of the water 
 down to 8 or less per cubic centimeter of the filtrate, and since 
 these very encouraging exhibits do not always occur simultane- 
 ously by months, it must be credited to management of the filters 
 or favorable company conditions, rather than to conditions prevail- 
 ing in the unfiltered waters. 
 
 It is apparent from the tables that the performance of the 
 filters of the London works for the winter months is in some 
 instances very unsatisfactory, and this must be due to a cause 
 which is susceptible of remedy. If it is chargeable to uncovered 
 filters, then covering should be resorted to. But the Hamburg 
 authorities assure the author that they have been able with un- 
 covered filters, and by an ingenious device for scraping the sand 
 under the ice-cake (see Fig. 26) which forms in their climate, to 
 keep the bacteria in the filtered water down to 30 per cubic cen- 
 timeter. It is well known that the winters are more rigorous 
 in Hamburg than in London ; and if it be possible to satisfy the 
 hygienic requirements in Hamburg during the winter, it surely 
 should be possible to do so in London. 
 
 It will also be noticed that the low bacterial counts in the 
 filtrate do not always follow the lower counts in the unfiltered 
 river water ; thus, the Chelsea filters were successful in reducing 
 the number of bacteria in the water to 4 per cubic centimeter 
 with 7,520, and to 2 per cubic centimeter with 4,300 in the ap- 
 plied water ; when with only 1,740 and 2,220 colonies in the river 
 water, the bacteria in the filtrate rose to 18 and 20 per cubic cen- 
 timeter. 
 
 For the West Middlesex Company the conditions seem to 
 change somewhat, the lower counts of bacteria in the river water 
 being followed by lower counts in the filtrate. This is not always 
 
150 THE PURIFICATION OF WATER. 
 
 true, even for this company : for with 26,800 bacteria per cubic 
 centimeter in the river water, the bacteria in the filtrate were as 
 low as 16 per cubic centimeter ; while with only 4,300 in the river 
 water, the bacteria in the filtrate were 14 per cubic centimeter of 
 the water. The cause of these variations and apparent vagaries 
 may be found in the relative condition of the filters in service 
 during the respective months. If the bacterial examinations hap- 
 pened shortly after one or more filters had been cleaned, the 
 counts in the filtrates might be relatively high ; while if the ex- 
 amination had just preceded the cleaning of a filter, the count 
 might be relatively low. 
 
 Aside from the poor work accomplished at times by the Lon- 
 don filters, it is manifest that filtration, properly conducted, can 
 produce remarkable changes in the bacterial contents and quality 
 of sewage-polluted waters. The general results as shown by the 
 tables indicate marvelous possibilities which can be attained by 
 filtration under a rigorous discipline upon the part of the health 
 authorities, and a proper accountability upon the part of the water 
 companies. 
 
 Failure to attain a low bacterial condition of the filtrate, as we 
 have seen, is not due to the weather, nor to the bacterial con- 
 tents of the unfiltered water, but to causes which an inflexible 
 regimen would speedily remove. A review of the seven water- 
 works of London which take their supply from rivers would be 
 unfair if it failed to state the fact that these companies were 
 hard pressed at times to secure the required quantity of water ; 
 and it is doubtless true, that if the supply of water from the 
 Thames and River Lea was at all times ample, and not a matter 
 of grave concern, the English engineers would undertake to sup- 
 ply a filtrate which should never exceed the prescribed standard 
 of bacterial contents, and usually be lower than that of average 
 spring waters. 
 
 The diagram on page 151 contains, to uniform scale, (1) the 
 depth or thickness of finer sand-bed ; (2) the rate in imperial 
 gallons per acre per day ; and (3) the percentage of bacteria in the 
 applied water removed by filtration, for each of the seven Lon- 
 don companies which filter their water supplies. The solid black 
 
FILTRATION OF WATER SUPPLIES. 
 
 151 
 
 surfaces show the thickness of sand-beds ; the finely hatched sur- 
 faces show the rate of nitration per acre per day ; and the coarsely 
 
 hatched surfaces the percentage of bacterial efficiency. The ordi- 
 nates originate at a common zero plane. 
 
152 
 
 THE PURIFICATION OF WATER. 
 
 The following very interesting table is taken from Mr. F. A. 
 Meyer's paper* on the Hamburg Water-Works (p. 23) : 
 
 NO. OF BACTERIA PER C. C. OF WATER AT VARIOUS POINTS OF THK 
 HAMBURG WATER-WORKS. 
 
 SOURCE OF SAMPLE. 
 
 BACTERIA PER C. C. 
 OF WATER. 
 
 Dec. 23, 1893 
 
 Jan. 17, 1894 
 
 New intake from River Elbe, Billwarder Island, 
 
 1,665 
 
 1,953 
 
 From settling-basins, 
 
 674 
 
 1,031 
 
 From main conduit to filters, " " 
 
 909 
 
 1,053 
 
 Unfiltered water from Filter No. 1, Kalte Hofe, 
 
 818 
 
 . . 
 
 " " " No. 3, " 
 
 
 1,094 
 
 " No. 19, 
 
 782 
 
 
 " No. 20, 
 
 . . 
 
 1,061 
 
 Filtered water from Filter No. 1, " 
 
 18 
 
 . . 
 
 " No. 2, 
 
 . . 
 
 33 
 
 " " " No. 3, " 
 
 . . 
 
 31 
 
 " " " No. 4, " 
 
 8 
 
 . . 
 
 " " " No. 5, " 
 
 18 
 
 . . 
 
 " " " No. 6, " 
 
 . 
 
 21 
 
 No> 7^ 
 
 7 
 
 
 " " " No. 8, " 
 
 . . 
 
 35 
 
 " " " No. 11, 
 
 33 
 
 28 
 
 " " " No. 12, " 
 
 45 
 
 30 
 
 " No. 15, 
 
 . . 
 
 . . 
 
 " No. 16, 
 
 24 
 
 16 
 
 " " " No. 17, " 
 
 29 
 
 14 
 
 " No. 18, 
 
 7 
 
 4 
 
 " No. 19, 
 
 11 
 
 9 
 
 " ' " No. 20, " 
 
 . . 
 
 18 
 
 " No. 21, " 
 
 . . 
 
 
 " " " No. 22, t " 
 
 . . 
 
 . . 
 
 From the new main collecting channel, " 
 
 19 
 
 17 
 
 tt 1(J <t tt 
 
 23 
 
 26 
 
 From siphons to the clear-water basin, Rothenburgsort, 
 
 25 
 
 23 
 
 From clear-water basin, " 
 
 31 
 
 24 
 
 From pumping-well, 
 
 28 
 
 40 
 
 From tap in the Hygienic Institute, Hamburg, 
 
 97 
 
 85 
 
 From tap at No. 25 Gunther Street, " 
 
 94 
 
 69 
 
 NOTE. Where no numbers of bacteria are given, the filters were not in service on the day 
 the test was made. 
 
 * Chapter XIII. 
 
 f The plan of the works at Hamburg contemplates twenty-two filters, of which Filters 
 Nos. 9, 10, 13, and 14 are to be constructed. 
 
FILTRATION OF WATER SUPPLIES. 153 
 
 The average for the 10 filters in service f)ec. 23, 1893, was 
 20 bacteria per cubic centimeter, while for the 11 filters in ser- 
 vice Jan. 17, 1894, the average was 22 bacteria per cubic centi- 
 meter in the filtered water, a reduction of 98.8 to 98.9 per cent of 
 the bacteria in the river water. 
 
 The increase in numbers from the filters to the taps at the 
 Institute of Hygiene, and on Gunther Street in the city, is not 
 in accord with the experience at Lawrence, Mass.,* where the 
 numbers of bacteria are generally found to be less as the water 
 passes from the clear-well of the filter to the service taps in the 
 city. 
 
 LAWRENCE, MASS. CITY FILTER. 
 
 BACTERIA PER C. C. AVERAGES FOR 1895. 
 
 OUTLET OF 
 
 MERRIMAC EFFLUENT DISTRIBUTING TAP AT TAP AT 
 
 RIVER WATER. FROM FILTER. RESERVOIR. CITY HALL. EXPERIMENT STATION. 
 
 10,666 122 122 84 77 
 
 From daily bacterial examinations for the six months ending 
 April 30, 1896, the following averages are given : f 
 
 LAWRENCE, MASS. CITY FILTER. 
 
 BACTERIA PER C. C. OF WATER. 
 
 OUTLET OF 
 
 MERRIMAC EFFLUENT DISTRIBUTING TAP AT TAP AT 
 
 RIVER WATER. FROM FILTER. RESERVOIR. CITY HALL. EXPERIMENT STATION. 
 
 7,533 134 119 86 85 
 
 The Hamburg filters were started in operation five months 
 before the Lawrence filter, and the influence of the better water 
 on the bacteria in the street mains should be no less in the former 
 than in the latter city. Such information as has come to the 
 author on this subject indicates that within a period of three 
 months or less, the regimen of the purified water is fully estab- 
 lished in the street mains and service pipes, and neglecting the 
 influence of a possible higher temperature of the water in the ser- 
 vice pipes than in the clear-well of the filters, the former should 
 show fewer bacteria than the freshly filtered water. The Law- 
 
 * Twenty-seventh Annual Report Massachusetts State Board of Health, p. 573, et seq. 
 t Ibid., p. 581. 
 
154 THE PURIFICATION OF WATER. 
 
 rence experience is in accord with this theory, and the author's 
 investigations along this line generally indicate that with filtered 
 water in pipes, and therefore not exposed to light and air, nor to 
 increased temperature, the bacteria diminish in numbers per cubic 
 centimeter after the water leaves the filters. 
 
 DOUBLE FILTRATION. 
 
 During the investigation by the Royal Commission on Metro- 
 politan Water Supply, it was suggested by Dr. E. Frankland,* 
 that the quality of the London river waters might be improved by 
 a second or double filtration. At first sight this proposition looks 
 very favorable, and if the operation of a sand filter was the same 
 as that of a fine strainer it would doubtless improve the quality 
 of the water by passing it successively through two or more filters 
 of diminishing sizes of sand ; but considering the action of a filter 
 in the removal of bacteria from water as a biological process, then 
 double filtration might yield a poorer filtrate than single filtration. 
 
 The formation of the gelatinous film on the surface of the 
 sand, and the closing of the pores, is hastened by, and, in fact, is 
 partly due to the deposition of the suspended organic matter in 
 the water. In the process of double filtration this organic matter 
 would be almost wholly arrested upon the first filter, leaving a 
 small amount of material in the water from which to form the 
 film on the second filter, and the partially purified water would in 
 all probability pass through the latter with little improvement in 
 quality. A preliminary filtration through a thin bed of coarse 
 sand, which would intercept the larger and heavier suspended 
 matters, might be of advantage in some instances, but as a rule, 
 double filtration cannot be regarded as the proper remedy for 
 unsatisfactory single filtration. 
 
 The biologic work of a plain sand filter has been so well 
 established that any proposition which inferentially rejects this 
 theory of action should be very carefully considered before it is 
 adopted in practice. 
 
 The film at the surface of the sand-bed consists partly of the 
 
 * Minutes of Evidence, p. 482, et seq. 
 
FILTRATION OF WATER SUPPLIES. 155 
 
 intercepted suspended matters, and partly of the products of bac- 
 terial action on the suspended organic matter in the water, and 
 this film in due time becomes in itself a very fine strainer, retain- 
 ing thus at the top of the sand-bed the food for support of bacte- 
 rial life ; and double filtration, to be successful, must contemplate 
 a free passage of much of the food material and bacteria through 
 the sand-bed of the first filter, to form the film on the sand of the 
 second filter. 
 
 The author would gauge the performance of a filter by its prac- 
 tical results, and practical results are to be found in the influence 
 of the water upon the typhoid fever rates of the city supplied from 
 such filter. It matters not how much the bacteria are reduced in 
 percentage of the applied water, or how far nitrogenous organic mat- 
 ter has been advanced towards nitrous and nitric acids, if the ty- 
 phoid fever rates have not been lowered. But with a reduction of 
 the bacteria, and conversion of organic matter into harmless com- 
 pounds, there ought to be a reduction in the typhoid fever rates. 
 That it should be so none will deny. But since the reduction of 
 the typhoid fever rates is what we are aiming at, why not make 
 this the measure of efficiency, and let the standard of operation of 
 filters be based upon the percentage reduction of case and death 
 rates, or better still, let the contracts and performance be based 
 upon a given death rate from typhoid fever, which shall follow the 
 use of such filtered water? 
 
 This is the practical result which cities desire, and for which, 
 when men are found with sufficient confidence, combined with 
 knowledge, to pledge such results, the author believes they are 
 willing to pay. It is all very well to point to a reduction of 99.9 
 per cent of the bacterial contents, and to the great chemical changes 
 which have taken place in the water while passing through the 
 filter ; but these are only steps in the journey, the end being the 
 reduction of the typhoid rates. 
 
 If the register of vital statistics shows a reduction by the use 
 of filtered water from 40 to 8 per 100,000 of population, or a per- 
 centage reduction of 80 in the typhoid rates, what need we care 
 about the degree of bacterial and chemical changes ? Here we have 
 
156 THE PURIFICATION OF WATER. 
 
 something tangible, something that the people, as well as the bacte- 
 riologist and chemist, can grasp, and appreciate at its full value. 
 
 INFLUENCE OF STORAGE ON FILTERED WATERS. 
 
 With reference to the time which water may be carried in sto- 
 rage after filtration, this will depend to some extent upon the origi- 
 nal source of the water. Surface waters always exposed to light 
 will be less affected by the growths of flora and fauna than ground 
 waters. No water, whether pure or purified, will be so destitute of 
 organic matter and bacteria as to no longer furnish the material 
 for further algous and bacterial development. But if, among the 
 forms of bacterial life in stored waters, there are none of the patho- 
 genic species, and in the organic matter in process of destruction 
 there are no ptomains, no objection from the standpoint of hy- 
 giene can possibly be raised to the processes which may be going 
 on in stored water. 
 
 The experience abroad universally favors the covering of sto- 
 rage or service reservoirs for filtered water. Whether this precau- 
 tion is absolutely necessary in all cases has not been shown as the 
 result of carefully conducted investigations, and it is possible that 
 certain waters can be filtered and carried without injury to quality 
 in open reservoirs for many days. Iron, lime, sulphur, and other 
 salts in solution in waters, favor the growth of some species of 
 vegetable life ; and the nature of the dissolved mineral substances 
 in water doubtless has a strong influence on its quality, if carried 
 in storage after filtration. 
 
 It is stated in connection with the Anderson Iron Purifiers at 
 Paris, that the water from these is carried for nearly three weeks 
 in covered reservoirs without change in bacterial contents, color, 
 taste, or odor. Upon the contrary, the water from the Lawrence, 
 Mass., filter is pumped to an open distributing reservoir ; and the 
 only cause of complaint which the author has noticed in the 
 Annual Reports since the filter was started in service has been 
 the formation of ice on the reservoir and filter. The reservoir, 
 when filled to a depth of 25 feet, has a water surface 694 by 375 
 feet, contains 40,000,000 gallons, and represents about 13 days' 
 
FILTRATION OF WATER SUPP 
 
 157 
 
 ' 
 storage. It consists of two equal compartments, and the water is 
 
 delivered to either or both divisions of the reservoir at will. 
 
 The experience at Quincy, 111., with stored filtered water, is 
 very different from that at Lawrence, Mass. At the former place, 
 since filtration has been practiced, during the summer months 
 "there is a vegetable growth of a mossy character which some 
 seasons imparts a fishy or woody taste to the water and other 
 seasons does not." At Quincy mechanical filtration with alum 
 is resorted to, while Lawrence depends upon plain sand filtration. 
 (The author is not aware of any investigations upon the influence 
 of undecomposed alum in water on certain species of plant life, and 
 it is possible when these are made, they may show that plants 
 having a strong affinity for potash or sulphur, will flourish in water 
 purified with alum.) The reservoir at Quincy has an available 
 water depth of 22 feet, a water surface of 350 by 250 feet, and 
 contains 18,000,000 gallons. The average daily consumption dur- 
 ing 1896 was 1,261,900 gallons, and the reservoir contains at the 
 present time about 14 days' supply. 
 
 By Act of Parliament, 1855, the water companies of London 
 were required to cover all distributing reservoirs which received 
 filtered water within a radius of five miles from St. Paul's ; and 
 according to the statistics of the London works for 1893, the 
 aggregate capacity of these reservoirs represented about four- 
 fifths of the average daily consumption, or 20 hours' supply. At 
 Berlin, Hamburg, and other cities of continental Europe, which, 
 like London, use filtered waters, the distributing reservoirs are 
 of small relative dimensions, and not intended for storage, but to 
 compensate for the varying rates of consumption during the day. 
 
158 THE PURIFICATION OF WATER. 
 
 CHAPTER XI. 
 
 TYPES OP SAND FILTERS. 
 
 SAND FILTERS may be classed as of four types : 
 
 1. The so-called continuous filters of European practice, as 
 
 designed by Mr. James Simpson of the Chelsea and 
 Lambeth Water- Works. 
 
 2. The intermittent filter, as designed by Mr. Hiram F. Mills, 
 
 C. E., for the water-works of Lawrence, Mass. 
 
 3. Natural filtration into collecting galleries, as illustrated in 
 
 the water-works of Lyons and Angers, France, and in 
 some cities of this country. 
 
 4. Mechanical filters, in which the sand-bed is restored to its 
 
 normal condition by washing, without removal from the 
 filter. 
 
 In situations where artificial filters are to be covered, there will 
 be a saving in space by constructing the walls of masonry backed 
 by puddle, as shown by Fig. 8 ; but with open filters, such as 
 are used in London and Hamburg, certain advantages are to be 
 had from the sloped walls, as shown by Fig. 13 (Chapter XIIL). 
 Where filters are liable to be covered with an ice-cake, and the 
 method of sand cleaning under the ice as practiced at Hamburg is 
 adopted (Fig. 26), then a sloped inside wall is absolutely essential 
 to the proper operation of this apparatus. The continuous filter is 
 usually rectangular in plan ; but this is varied to suit the location, 
 some of the London filters being circular. So far as form affects 
 filtration, the beds may take any of the regular shapes in plan ; but 
 for convenience of scraping and removing the sand, the rectangular 
 form is to be preferred. Regulating devices to limit the head on 
 a filter are regarded now as essential features of such works, in 
 order to remove as far as possible from the attendants the power 
 
TYPES OF SAND FILTERS. 
 
 159 
 
 to seriously interfere with the regularity of operation of a filter. 
 These regulating devices generally consist of floats which rise and 
 
 Section of Gathering Drain. 
 
 Fig. 8. St. Louis Water-Works. Details of Filters. 
 
 fall with the water in the inlet and outlet wells, actuating large 
 balanced valves which are automatically raised and lowered to 
 admit more or less water through the passages. 
 
160 THE PURIFICATION OF WATER. 
 
 In cases where automatic regulators are lacking, sliding weirs 
 and valves or gates, adjustable by hand, under intelligent super- 
 vision, will accomplish the same results. In order that the effective 
 head on the filter and rate of nitration might be conveniently reg- 
 ulated, Mr. Kirkwood proposed the device shown by Fig. 8, con- 
 sisting of a weir which could be raised or lowered in guides by 
 means of the winch shown at the ground level. With this device 
 the draught upon the clear- well, and the raising and lowering of 
 the water-level therein, can have no effect upon the rate of delivery 
 of the filter. 
 
 The continuous sand filter of European practice consists of a 
 shallow reservoir with inclined or sloping sides when made with 
 earthen walls, or with vertical sides if made of masonry.* The 
 surface area of these reservoirs ranges from an acre or less to as 
 much as two acres. The depth of the reservoir is 10 or 12 feet, 
 according to the local character of the water, the materials avail- 
 able for the filter-beds, and the views of the designer. 
 
 The two general methods of construction of the basin part of 
 plain sand filters is clearly shown by Fig. 8, which is a section 
 of the filter proposed by Mr. Kirkwood for the city of St. Louis 
 (1866), and by Fig. 15 (Chapter XIII.), which is a plan and sec- 
 tion of the filters of Hamburg. When the walls are vertical, as 
 shown by Mr. Kirkwood's plan, these must be of masonry, backed 
 with puddle or an impervious clay. When made with slopes inside, 
 the slope should be of water-tight materials, or puddled, and faced 
 with a pavement or lining of concrete, asphalt, brick, or stone set 
 in mortar. Should ice form on the water in the filter, the thrust 
 of the cake is less liable to injure the sloped sides than the vertical 
 walls, while the cost of construction favors the basin with sloped 
 walls. 
 
 In the bottom of this reservoir is placed a system of lateral 
 parallel drains, which collect the water from all parts of the filter, 
 and conduct it to one or more main central drains, by which it is 
 removed from the filter to the clear-well, and from this it is pumped 
 to distributing reservoirs, or direct into the distributing mains. 
 
 Over the drains in the bottom of the filter is laid a course of 
 
 * " Hygiene of Water," by the author, Dietetic and Hygienic Gazette, 1896, p. 599. 
 
TYPES OF SAND FILTERS. 161 
 
 
 
 broken stone or large gravel about 12 or 15 inches thick ; over 
 this a layer of small gravel 6 inches thick ; over this a layer of 
 coarse sand 6 to 12 inches thick; and on top of the coarse sand is 
 placed a layer of graded sand from 24 to 36 inches thick. This 
 upper layer of sand is the real filtering material, the layers of 
 sand and gravel below simply being beds for the support of this 
 finer bed of graded sand at the top of the filter. 
 
 The capacity of a filter is stated as the average number of 
 gallons which it will deliver per acre per day during the time it is 
 in service. When the filter is new and the sand surface clean, the 
 rate of delivery will be quite up to the average rate, even with a 
 few inches difference of water level over the sand and in the efflu- 
 ent chamber ; but as the sand becomes clogged with the suspended 
 matters in the water, and by the products of bacterial action upon 
 the organic matter, the effective head of water on the sand-bed 
 must be gradually increased, to filter the water at or near the 
 standard rate, until in due time the limiting head is reached, and 
 the rate of delivery of the filter diminishes until it is no longer 
 profitable to operate it, when the water is shut off from it altogether, 
 and the filter taken temporarily out of service. 
 
 The upper layer of graded sand is then carefully scraped for a 
 depth of about one-half inch, the dirty sand removed and washed 
 by mechanical apparatus devised for the purpose, and stored in the 
 sand-house preparatory to putting it back on the filter whenever 
 the successive scrapings have reduced the thickness of the upper 
 sand-bed to 15 or 20 inches. The thickness of the upper layer of 
 sand is seldom reduced to less than 15 inches, and the lower layers 
 of sand and gravel are not disturbed or renewed at all unless some 
 radical overhauling of a filter is shown to be necessary by the bac- 
 terial analyses of the water. 
 
 During the time of service of such a filter, it is acting continu- 
 ously without interruption, day and night ; and the rate of delivery 
 is maintained either by manual labor or automatic regulating de- 
 vices as near to an established standard as possible. 
 
 The rate of delivery will depend (1) upon the grade of sand 
 in the upper layer, (2) the effective head of water on the sand-bed, 
 and (3) the condition of the water when it comes to the filter. 
 
162 THE PURIFICATION OF 'WATER. 
 
 Neglecting the grade of sand and head, which are subject to con- 
 trol, it can generally be said that the rate of operation will depend 
 upon the condition of the water. Therefore, when filters are 
 operated in connection with subsiding reservoirs, i.e., where the 
 work of purification is a process of filtration combined with sedi- 
 mentation, the larger the amount of work done in the subsiding 
 reservoirs, the finer may be the grade of sand used, and the higher 
 the rates of delivery of the filters per unit of sand surface. The 
 smaller size of sand-grain within certain limits will insure an im- 
 proved quality of filtrate, and the higher rate of delivery will effect 
 a reduced cost of treatment per million gallons of water. 
 
 Grading of Filter Materials. In order that a filter shall give 
 satisfactory results, both in rate of operation and quality of filtrate, 
 it is necessary that the filtering materials, especially the bed of sand, 
 shall be selected with regard to the size of its grains, and the rela- 
 tion to each other by weight of these several sizes in a bed of mixed 
 sand. In practice, wherever much experience has been had with 
 sand filters, the materials are classified as boulders (or broken 
 stone), gravel, and sand, the gravel being used frequently of two 
 sizes, and denominated as coarse and fine gravel, and the sand of 
 two general grades, a coarser grade above the small gravel, and 
 the finer filtering-sand at the top of the filter-bed. 
 
 Filtering materials, as used at the experiment station of the 
 Massachusetts State Board of Health, are examined as to " size " 
 and " uniformity " of size of grains or particles, and graded to suit 
 the particular work to which they are applied. Chemical and bac- 
 terial analyses are also made of the matter attached to the sand 
 grains, which it appears, even after repeated washings, is not 
 entirely removed.* 
 
 The larger materials can readily be graded by hand picking ; 
 sands not less than 0.10 mm. mean diameter are separated and 
 graded into commercial sizes by a series of sieves of brass wire, set 
 in metal rims and shaken in a machine, for a time sufficient to 
 secure the passage through the sieves of all but the particles larger 
 than the mesh of the wire cloth, while particles of sand smaller 
 than 0.10 mm. in any diameter are graded by water elutriation. 
 
 * Miiggel Lake Water Supply, by Henry Gill, p. 11. 
 
TYPES OF SAND FILTERS. 163 
 
 The method of separation of a mixed sand* into its several sizes 
 by means of sieves is obvious, and requires no explanation. Water 
 elutriation of sand consists in adding to a volume of distilled 
 water, measured in a beaker, a definite weight of clean, dry sand ; 
 and after a thorough mixing of the sand and water by means of a 
 strong current of air passed through a glass tube, a given time is 
 allowed for the precipitation by gravity of the sand to the bottom 
 of the beaker. It is well known that after the mixing of solids in 
 a fluid, the larger and heavier particles settle first, and an experi- 
 ment will demonstrate that all grains of sand of not less than a 
 given diameter will be precipitated within a given time. By allow- 
 ing more time, grains of smaller size and weight will settle out of 
 the water ; and by allowing sufficient time, all the grains, however 
 small they may be, will have been precipitated to the bottom of 
 the beaker, and the distilled water above will be free from all sus- 
 pended sand. 
 
 Adopting the results of sand measurement by the Massachu- 
 setts State Board of Health (with 230 cubic centimeters of water 
 and 5 grams of sand), after a thorough mixing of the sand in the 
 water, all grains which are precipitated to the bottom of the beaker 
 within 15 seconds are considered as of not less than 0.08 mm. diam- 
 eter. Similarly, all grains, which upon mixing and allowing one 
 minute for subsidence are collected at the bottom of the beaker, 
 are regarded as of not less than 0.04 mm. diameter, and all grains 
 which remain suspended in the water at the end of one minute are 
 regarded as 'of less than 0.04 mm. diameter. The weight of the 
 smallest particles in a mixture of sand, the largest grains of which 
 are less than 0.10 mm. diameter, is obtained by deducting from 
 the whole mass (5 grams) the added weight of the grains larger 
 than 0.08 mm. diameter, and the grains less than 0.08 mm. but 
 larger than 0.04 mm. diameter. The difference is held to be the 
 weight of the grains less in diameter than 0.04 mm. ; i.e., sizes from 
 0.10 mm. down to 0.08 mm., and from 0.08 mm. to 0.04 mm., are 
 determined by water elutriation, while smaller sizes are held to 
 be the difference between the whole weight and the sum of the 
 weights of the two larger sizes. 
 
 In stating the dimensions used for comparison of sand-grains 
 
164 THE PURIFICATION OF WATER. 
 
 in a mass of sand constituting a filter-bed, two terms have been 
 proposed by the Massachusetts State Board of Health* which are 
 convenient for general use. 
 
 Effective Size of Sand-Grain. This is that diameter of grain 
 in a mass of sand of which 10 per cent of the mass by weight is 
 smaller, and 90 per cent is larger in size. Thus, if upon physical 
 analysis of a body of sand by sieves or any convenient method, it 
 is found that the largest diameter for the 10 per cent by weight of 
 the smaller grains is 0.50 mm., and 90 per cent is of diameter 
 larger than 0.50 mm., then 0.50 mm. would be regarded as the 
 "effective size." Or, assuming the mass of sand to be regularly 
 graded from the finest to the coarsest particles, then the " effec- 
 tive size" will be that size of which 10 per cent of the whole mass 
 by weight is smaller in diameter. 
 
 Uniformity Coefficient. This is the ratio of the diameter of 
 the sand-grain of which by weight 60 per cent is finer than itself, 
 to the diameter or size of which 10 per cent is finer than itself. 
 If by weight 60 per cent of a sample of sand is less than 0.50 mm., 
 and 10 per cent is less than 0.25 mm., the " uniformity coefficient " 
 is =g = 2 ; or the diameter of grains of which 60 per cent is finer 
 than itself, divided by the "effective size," gives the " uniformity 
 coefficient." 
 
 The lower the "uniformity coefficient," i.e., the more regular 
 in size are the grains of sand in any mass, the larger will be 
 the water space or voids. With a high " uniformity coefficient," 
 i.e., with great irregularity of size in the sand-grains, the smaller 
 will be the water space or voids. 
 
 The influence on the bacterial efficiency of filters of the " effec- 
 tive size " of sand-grains, with widely different rates of filtration, 
 is shown by the table on the following page. 
 
 The term " effective size " must not be confounded with the 
 average size of sand-grain in a sample. The average size in a 
 mass of mixed sand being always larger in diameter than the 
 " effective size." The results obtained at Lawrence indicate that 
 the finer 10 per cent of sand has as much influence on filtration 
 as the coarser 90 per cent. 
 
 * Twenty-fourth Annual Report Massachusetts State Board of Health, p. 541. 
 
TYPES OF SAND FILTERS. 
 
 165 
 
 EXPERIMENTAL FILTERS, LAWRENCE, MASS., 1895. 
 
 
 RATE OF 
 
 EFFECTIVE SIZE 
 
 
 
 TYPE OF FILTER. 
 
 FILTRATION. 
 GALLONS PER ACRE 
 
 OF 
 
 SAND-GRAINS. 
 
 UNIFORMITY 
 COEFFICIENT. 
 
 BACTERIAL 
 EFFICIENCY. 
 
 
 PER DAY. 
 
 MILLIMETERS. 
 
 
 
 Continuous, 
 
 1,980,000 
 
 0.14 
 
 2.2 
 
 99.49 
 
 
 
 3,576,000 
 
 0.23 
 
 2.3 
 
 99.73 
 
 ( 
 
 6,780,000 
 
 0.29 
 
 2.7 
 
 99.37 
 
 (i 
 
 4,280,000 
 
 0.38 
 
 3.5 
 
 99.51 
 
 (C 
 
 4,680,000 
 
 0.48 
 
 2.4 
 
 99.45 
 
 Intermittent, 
 
 1,960,000 
 
 0.14 
 
 2.2 
 
 99.75 
 
 (t 
 
 3,096,000 
 
 0.23 
 
 2.3 
 
 99.16 
 
 tt 
 
 6,600,000 
 
 0.29 
 
 2.4 
 
 99.38 
 
 
 
 4,500,000 
 
 0.48 
 
 2.4 
 
 99.57 
 
 Sterilization of Filter Sand. The sterilization of the sand-bed 
 of the mechanical filter used in the Providence, R.I., tests* was 
 shown by Mr. Weston to have no influence upon the bacterial 
 efficiency of the filter ; boiling of the sand in water for " one hour 
 and fifty minutes " did not improve its capacity to restrain the 
 passage of bacteria. Piefke, of the Berlin Water- Works,f previ- 
 ously had experimented with sterilized sand in a small filter, with 
 the result that "more organisms were found in the filtrate than 
 in the unfiltered water." 
 
 The same result has repeatedly been shown during the experi- 
 ments of the Massachusetts State Board of Health at Lawrence.J 
 " Heating the sand and pouring the boiling water through it 
 caused 100 times as many bacteria to pass through it with 
 sewage for three months as passed through a similar filter whose 
 sand and first water had not been heated." 
 
 Experience has amply demonstrated that the proper treatment 
 of sand for use in water filters is a thorough washing, finally, with 
 filtered water, until all detachable matter is removed from the 
 grains, and in this condition the sand placed in the filter. The 
 improved chemical and bacterial results of washed over sterilized 
 sand suggests that the bacteria, which find a suitable nidus in a 
 
 * Report on Results Obtained -with Experimental Filters, Providence, R.I., p. 119. 
 
 t 1887. 
 
 \ H. F. Mills, Transactions American Society Civil Engineers, vol. xxx., p. 359. 
 
166 THE PURIFICATION OF WATER. 
 
 sand-bed after the sand has been washed, are useful in the destruc- 
 tion of organic matter and of some species of water bacteria. 
 
 LAWRENCE, MASS., INTERMITTENT SAND FILTER. 
 
 This filter is the outgrowth of two forces : (1) the abnormally 
 high typhoid fever rates of the city of Lawrence ; and (2) the 
 labors of the State Board of Health along the line of water purifi- 
 cation. It consists of a single filter with a sand area of 2.50 acres ; 
 and instead of a horizontal surface, like the European filters, the 
 bed of sand is furrowed or channeled from side to side to provide 
 a uniform thickness of sand through which the water from all 
 points is compelled to percolate to the collecting drains at the 
 bottom of the filter. 
 
 The mean elevation of the surface of the sand-bed is 2 feet 
 below low water in the Merrimac River, and the filter is flooded for 
 this depth from the river. The tank in which the filter is con- 
 structed consists of an excavation in the bank of the river, with a 
 bottom elevation averaging 7 feet below low-water mark. 
 
 The general dimensions of this filter, as given in the Report 
 of the Massachusetts State Board of Health for 1893, are as 
 
 follows : 
 
 Width of filter, 150 feet. 
 
 Length of filter, 750 " 
 
 Effective area of sand surface, 2.50 acres. 
 
 Depth of filtering-sand, 5.00 feet. 
 
 Effective size of sand, 0.25 mm. 
 
 The underdrains were placed 30 feet apart. On the line of 
 each drain the excavation is carried down to elevation 8 feet below 
 low-water mark in the river, for a width of 5 feet ; and the crests 
 of the ridges midway between the underdrains are at elevation 
 6 feet below low-water mark, and 5 feet wide. From the crest 
 down to the channel for the underdrains, the bottom is sloped at 
 the rate of 5 feet horizontal to 1 foot vertical. 
 
 Over the ridges the depth of filtering-sand was made 3 feet 
 for a width of 5 feet, and over the line of underdrains the depth 
 of sand was made 6 feet for a width of 5 feet, the surface 
 
TYPES OF SAND FILTERS. 
 
 167 
 
 slope of the sand from the ridges to 
 the crests being 10 feet horizontal to 
 1 foot vertical. 
 
 The section given in Fig. 9 indi- 
 cates the manner in which the bottom 
 of filter and surface of sand-bed were 
 furrowed. 
 
 From the description in the report, 
 it appears that each furrow or channel 
 with its underdrain and superimposed 
 filtering material is complete in itself, 
 the underdrain terminating in a 10-inch 
 pipe which passes through the masonry 
 wall of an old filter gallery (lying paral- 
 lel to the new filter). This filter gal- 
 lery became, upon completion of the 
 new intermittent filter, the clear-water 
 reservoir from which the pumps take 
 water. Assuming each 30 feet, then, a 
 complete section of the filter, there are 
 25 such sections in the whole bed. 
 
 The underdrains consist partly of 
 broken stone, and partly of glazed 
 sewer pipe from 4 to 10 inches diam- 
 eter, the larger pipe going through the 
 wall of the former filter gallery. Con- 
 sidering an underdrain 150 feet long 
 (width of filter), the broken stone and 
 pipe are placed as follows : 
 
 Broken stone, 90 feet. 
 
 Four-inch sewer pipe, 6 " 
 
 Six-inch sewer pipe, 50 " 
 
 Eight-inch sewer pipe, 4 
 
 Total, 150 feet. 
 
 The general slope or grade of the 
 drain is given as 1 foot fall in 100 feet 
 of length. 
 
 :-*-- 
 
 '-* 
 
 ',9 
 
 Fig. 9. 
 
 Longitudinal Section of Lawrence 
 (Mass.) Filter. 
 
168 THE PURIFICATION OF WATER. 
 
 The water is brought from the river through a 24-inch iron 
 pipe, which discharges into an open conduit, at one side of the 
 filter, from which channels of concrete are extended to within 32 
 feet of the farther side of the sand-bed. This conduit, together 
 with the lateral channels, are intended to secure a uniform distri- 
 bution of the water over the surface of the filter bed. 
 
 In the construction of this filter, the bottom (as is customary 
 to prevent entrance and mixing of ground water with the filtered 
 water) was not made water-tight, neither was the filter covered as 
 the climate seems to require, for the reason as given by Mr. Hazen.* 
 
 "It was no easy matter to secure the consent of the city government to 
 the expenditure of even the sum used ; there was much skepticism as to the 
 process of filtration in general, and it was said that mechanical niters could be 
 put in for about the same cost. Insisting upon the more complete and expen- 
 sive form might have resulted either in an indefinite postponement of action, 
 or in the adoption of an inferior and entirely inadequate process. Still, I feel 
 strongly that in the end the greater expense would have proved an excellent 
 investment in securing softer water, and in the greater facility and security of 
 operating the filter in winter." 
 
 The filter was proportioned for a daily rate of 2,000,000 gallons 
 per acre, assuming that it would be in service for 16 hours and at 
 rest for 8 hours, which would make the actual rate 3,000,000 gallons 
 per acre per day while in service. According to the report for 
 1895, the rate was 1,200,000 gallons per acre per day, as an average 
 for the year. 
 
 During the year 1895, 1,096,000,000 gallons of water were 
 passed through the filter, f and the total expense for maintenance 
 was 87,400, or $6.75 per million gallons of water filtered. 
 
 For scraping and replacing the sand, 
 For removing ice, 
 
 For washing 1,500 cubic yards of sand, 
 Total, 
 (Laborers paid $2.00 per day.) 
 
 Omitting cost of removing ice, which would be unnecessary in 
 milder climates, the cost per million gallons of water filtered be- 
 
 * Filtration of Public Water Supplies, New York, 1895, p. 98. 
 
 f Twenty seventh Annual Report Massachusetts State Board of Health, p. 572 et seq. 
 
TYPES OF SAND FILTERS. 169 
 
 comes 84.10. This being a single filter on the intermittent plan, 
 it cannot be handled so advantageously as filters in series ; and the 
 cost of operation per million gallons should be greater than in a 
 large plant embracing a series of continuous filters, in which one 
 or more filters could be taken out of service from time to time and 
 cleaned at convenience. 
 
 The filter was operated for 1895 at an average rate of -WV- = 
 3,000,000 gallons per day. With an effective area of sand surface 
 of 2.5 acres, and assuming the filter to have been in service two- 
 thirds of the time, the average rate of filtration per acre was 
 1,800,000 gallons per day. Estimating the population of Law- 
 rence for 1895 at 50,000, the average daily consumption of water 
 amounted to 60 gallons per capita. 
 
 The following are the average bacterial results for the year : 
 
 Bacteria in Merrimac River water, 5,000- 20,000 colonies per c. c. 
 
 Bacteria in effluent from filter, 38- 368 colonies per c. c. 
 
 Average bacterial reduction, 98.4 per cent. 
 
 This filter is reported to have cost 65,000, or $26,000 per 
 acre of effective filtering surface. 
 
 In speaking of the merits of plain sand filtration, Mr. J. Her- 
 bert Shedd, C.E., who has given very earnest consideration to 
 the matter of improvement of the Providence, R.I., water supply, 
 says : * 
 
 " The construction of sand filters for slow filtration is not an experiment. 
 Their use and value have been demonstrated in great numbers of cases through 
 long periods of years. The only reasons that I know for departing from this 
 long-established practice is in the effort to get cheaper first cost, or because 
 the necessary area for their establishment is not available. In the case of the 
 city of Providence there are about 70 acres of suitable land near the pumping- 
 station, bought for this purpose more than 20 years ago and still available. 
 Since it has been found that a mechanical filter is likely to cost $281,000, 
 plans have been made, levels taken, test pits dug, and estimates made for a 
 sand filter for slow filtration on that land, with all necessary settling space, 
 supply conduits and screens, leading mains, and other appurtenances, complete 
 for connecting the plant with our present works. My estimate of the cost of 
 
 * Engineering Record, July 22, 1894. 
 
170 
 
 THE PURIFICATION OF WATER. 
 
 supplying all materials and doing this work is about $208,000 ; but whether 
 my estimates have any value or not, the commissioner of public works has in 
 his possession a bid from responsible contractors who offer to do the whole 
 work complete for $200,000'. 
 
 " The cost of annual maintenance of the slow filtration plant will be, I 
 think, very materially less than the cost of maintaining a mechanical plant of 
 the same daily capacity." 
 
 PLAIN SAND FILTRATION. 
 
 The Report on the Providence experiments furnishes some 
 interesting information upon plain sand filtration with a filter of 
 
 Fig. 10. Experimental Filter, Providence, R. I. 
 
 <t 
 
 the form shown in Fig. 10. Two filters of this style were tested 
 by Mr. Weston, from March 27 to Oct. 5, 1893, with occasional 
 
TYPES OF SAND FILTERS. 
 
 171 
 
 intermissions for cleaning filters and changing sand ; and for tests 
 of the influence of alum, on so-called natural filtration. These 
 filters were worked at rates of 1,000,000 to 35,000,000 gallons 
 per acre per day. (Fig. 10.) 
 
 Considering Filter No. 1, from May 15 (before which date the 
 operation was discontinuous) to July 15 (after which date the con- 
 ditions of operation were varied almost daily), the rates of filtra- 
 tion varied from 1,000,000 to 3,500,000 gallons per acre per day. 
 Considering Filter No. 2, from April 15 (before which date the 
 operation was quite irregular, doubtless due to the starting of 
 novel experiments) to Sept. 13 (after which date the operation 
 was discontinuous), the rates of filtration varied from 1,000,000 
 to 35,000,000 gallons per acre per day. In the review of the per- 
 formance of Filter No. 2, the work is omitted from July 25 to 
 July 29 (four days), when alum was used. These filters were 
 restored to service after the sand-beds became clogged, sometimes 
 by washing the sand with a reverse current of water, and some- 
 times by scraping the sand. The following table will show how 
 this occurred by dates for the two filters : 
 
 DATE. 
 
 FILTER No. 1. 
 
 PERIOD OF 
 SERVICE, DAYS. 
 
 FILTER No. 2. 
 
 PERIOD OF 
 SERVICE, DAYS. 
 
 1893. 
 
 
 
 
 
 March 27, 
 
 Started, 
 
 . . . 
 
 Started, 
 
 . . . 
 
 April 21, 
 
 Sand renewed, 
 
 19 
 
 . . . 
 
 . . . 
 
 " 29, 
 
 Sand washed, 
 
 3 
 
 . . . 
 
 ... 
 
 May 10, 
 
 Sand renewed, 
 
 4 
 
 . . . 
 
 . . . 
 
 " 11, 
 
 
 . . . 
 
 Sand repacked, 
 
 33 
 
 June 14, 
 
 Sand scraped, 
 
 26 
 
 . . . 
 
 . . . 
 
 July 14, 
 
 . . . 
 
 ... 
 
 Sand washed, 
 
 51 
 
 " 16,* 
 
 Sand scraped, 
 
 26 
 
 . . . 
 
 . . . 
 
 " 24, 
 
 . . . 
 
 
 Sand washed, 
 
 8 
 
 " 27, 
 
 . . . 
 
 . . . 
 
 Sand washed, 
 
 3 
 
 August 23,* 
 
 . . . 
 
 . . . 
 
 Sand scraped, 
 
 21 
 
 When the filters were cleaned by washing, the current of water 
 was reversed as in mechanical filters. At times the sand-beds 
 were scraped in the usual way, taking off about one-half inch of 
 
 * After these dates the manner of operation was very irregular. 
 
172 
 
 THE PURIFICATION OF WATER. 
 
 sand. The size of sand used in these niters is given in the 
 Report as follows : 
 
 DATE. 
 
 FILTER No. 1. 
 
 FILTER No. 2. 
 
 March 27, 
 
 Effective size 0.81 mm., uniform- 
 
 Effective size 0.81 mm., uniform- 
 
 
 ity coefficient, 2.2. 
 
 ity coefficient, 2.2. 
 
 April 21, 
 
 Effective size 0.18 mm., uniform- 
 
 
 
 ity coefficient, 2.1. 
 
 
 May 10, 
 
 Effective size 0.35 mm., uniform- 
 
 Repacked bed with same sand. 
 
 
 ity coefficient, 2.0. 
 
 
 July 29, 
 
 Repacked bed with same size 
 
 
 
 sand. 
 
 
 Aug. 14, 
 
 Effective size 0.81 mm., uniform- 
 
 
 
 ity coefficient, 2.2. 
 
 
 Considering these niters after April 15, when they were started 
 in service at regular and standard rates of filtration, Filter No. 1 
 was in operation for 63 days, of which 36 days gave bacterial effi- 
 ciencies for the effluent of over 98 per cent, and 22 days gave 
 bacterial efficiencies of over 99 per cent, with bacterial counts in 
 the filtrate at times as low as 2, 4, 8, and 10 colonies per cubic cen- 
 timeter. For the same interval of time, Filter No. 2 was ope- 
 rated for 69 days, of which 41 days gave bacterial efficiencies of 
 more than 98 per cent, and 30 days gave bacterial efficiencies 
 of more than 99 per cent, with bacterial counts in the nitrate as 
 low as 6, 9, 10, and 12 colonies per cubic centimeter. 
 
 From July 17 to July 24, a period of seven days, Filter No. 2 
 was operated at an average rate of 29,043,000 gallons per acre per 
 day, with an average bacterial efficiency of 98.5 per cent. Dur- 
 ing this period the highest bacterial efficiency was 99.9 per cent, 
 the bacteria being reduced from 9,067 per cubic centimeter in the 
 applied water to 10 per cubic centimeter in the filtered water, and 
 the lowest bacterial efficiency was 92.9 per cent, when the bacteria 
 was reduced from 565 per cubic centimeter in the applied water to 
 40 per cubic centimeter in the filtered water. 
 
 The average bacterial efficiency for the Morison Mechanical 
 filter, using alum, from July 20, 1893, to Jan. 30, 1894,* is given 
 as follows : 
 
 * Report on Results Obtained -with Experimental Filters, Providence, R.I., 1896, pp. 55-57. 
 
TYPES OF SAND FILTERS. 173 
 
 
 
 For end growths, 97.21 
 
 For ninety-hour growths," 97.86 
 
 From Aug. 15 to Sept. 13 (1893), Filter No. 1 was operated 
 with "alum and free flow" for a period of 23 days, while Filter 
 No. 2 was operated as a plain sand filter for 22 days. During 
 this interval of time both filters were packed with sand-grains of 
 an " effective size" of 0.81 mm., and a "uniformity coefficient" of 
 2.2. The rates of filtration and bacterial efficiencies are tabled 
 below : 
 
 FILTER RATE OF FILTRATION. BACTERIA. 
 
 GALLONS PER ACRE PER DAY. PERCENTAGE REMOVED. 
 
 No. 1 (with alum), 30,100,000 82.11 
 
 No. 2 (without alum), 30,900,000 89.45 
 
 These figures indicate no advantage in the use of alum, not- 
 withstanding the rate of filtration was less than one-fourth of the 
 standard rate which Mr. Weston has proposed for filters of the 
 mechanical class using alum as a coagulant. Filter No. 2 was 
 always operated with an "effective size" of sand-grains of 0.81 
 mm., while Filter No. 1 was operated with sand-grains of an 
 "effective size" of 0.81 mm., 0.18 mm., and 0.35 mm. 
 
 During the interval of time taken for comparison of filtration 
 "with alum" and filtration "without alum," the "effective size" 
 of sand-grains, condition of the applied water, and rates of filtra- 
 tion were the same. The "effective size" of sand-grains for the 
 Morison mechanical filter was 0.59 mm., with a "uniformity 
 coefficient " of 1.5. 
 
 The influence of sand finer in "effective size" than 0.81 mm. 
 is shown by comparison of the operation of Filter No. 1, when 
 it was packed with sand of an "effective size" of grain 0.35 mm., 
 with Filter No. 2, which was packed with sand-grains of the larger 
 size. 
 
 The average bacterial efficiencies of Filter No. 1, operating 
 without alum, from May 25 to July 15, 1893, a period of 42 
 days, during which time the sand-bed was scraped once, were as 
 follows : 
 
174 THE PURIFICATION OF WATER. 
 
 FILTER No. 1. Sand-Bed Scraped and Filter started in service May 15. 
 
 EFFECTIVE SIZE OF SAND, 0.35 MM. UNIFORMITY COEFFICIENT, 2.0. 
 
 AVERAGES. 
 
 _ RATE OF FILTRATION. PERCENTAGE OF 
 
 INTERVAL OF TIME. _ _ 
 
 GALLONS PER ACRE PER DAY. BACTERIA REMOVED. 
 
 May 25-June 13, 2,300,000 98.3 
 
 During above test the numbers of bacteria per cubic centimeter 
 in the effluent were often below 100, and at times as low as 8, 24, 
 and 26. 
 
 Sand-Bed Scraped and Filter Started in Service June 15. 
 
 AVERAGES. 
 
 RATE OF FILTRATION. PERCENTAGE OF 
 
 GALLONS PER ACRE PER DAY. BACTERIA REMOVED. 
 
 June 15-July 15 2,280,000 97.9 
 
 During this latter interval the numbers of bacteria per cubic 
 centimeter in the filtrate were often less than 100, and at times 
 as low as 2, 4, 9, 10, 11, 13, and 18. 
 
 The average bacterial efficiencies of Filter No. 2, operating 
 without alum, from May 25 to July 13, 1893, a period of 40 days, 
 during which time there was no scraping of the sand-bed, were as 
 follows : 
 
 FILTER No. 2. Sand Removed and Filter Started in service May 13. 
 
 EFFECTIVE SIZE OF SAND, 0.81 MM. UNIFORMITY COEFFICIENT, 2.2. 
 AVERAGES. 
 
 RATE OF FILTRATION. PERCENTAGE OF 
 
 GALLONS PER ACRE PER DAY. BACTERIA REMOVED. 
 
 May 25-June 13 2,316,400 98.1. 
 
 During this interval of time the bacteria in the effluent were as 
 low as 14 to 16 per cubic centimeter. 
 
 Sand-Bed Unscraped Since Previous Use. 
 
 AVERAGES. 
 T RATE OF FILTRATION. PERCENTAGE OF 
 
 GALLONS PER ACRE PER DAY. BACTERIA REMOVED. 
 
 June 15-July 13 2,296,250 97.4 
 
 During this interval of time the numbers of bacteria in the 
 effluent were as low as 9, 12, and 16 per cubic centimeter. 
 
 A comparison of these tables indicates no special advantage in 
 the sand of smaller size of grain. 
 
TYPES OF SAND FILTERS. 175 
 
 
 
 These results taken as a whole clearly show that plain sand 
 filtration as conducted during the Providence, R.I., experiments is 
 quite as efficient as mechanical filtration with " alum," and is not 
 calculated to impart an astringency or acidity to the filtrate, which 
 may be positively hurtful to some systems. 
 
 Mr. Weston's experiments with plain sand filters in the author's 
 opinion are entitled to more consideration than they have been 
 given in the official Report. Here are two sand filters, 30 inches 
 diameter, operating naturally at rates 1,000,000 to 35,000,000 gal- 
 lons per acre per day, with average bacterial efficiencies of 97, 98, 
 99, and occasionally 100 per cent (the higher results sometimes 
 being obtained with the higher rates of filtration)'. The average 
 efficiency for Filter No. 2, without alum, at rates of 25,000,000 to 
 35,000,000 gallons per acre per day for about the same length of 
 time, is the same as the average bacterial efficiency of the Mori- 
 son mechanical filter after the sand-bed was washed with a solu- 
 tion of caustic soda. 
 
 LOWELL, MASS., FILTER-BED. 
 
 The following description of this filter is taken from the Manual 
 of American Water-Works, 1889-1890, p. 63. Population, 78,000. 
 
 Built in 1876, in gravel between the filter gallery and the 
 river : 100 by 114 feet at bottom, which is 8 feet below the level 
 of the Pawtucket dam. On the gravel is laid a dry stone drain, 
 15 inches square at the river end, 100 feet long, and 30 inches 
 square at the end nearest the gallery, where it terminates in a 10- 
 foot circular brick chamber, connected with the filter gallery by a 
 30-inch pipe. From the central drain 27 lateral stone drains are 
 laid, each 8 by 12 inches. The filtering materials consist of: 
 
 18 inches fine sand. 
 
 6 " coarse screened sand. 
 10 " coarse gravel (J-inch diameter). 
 36 " -inch cobble stones. 
 70 inches total depth of filtering materials. 
 
 " In the spring of 1877 a freshet deposited 18 inches of sand 
 and silt on the filter-bed ; and it was necessary to admit some water 
 
176 THE PURIFICATION OF WATER. 
 
 directly from the river until the deposit was removed in Septem- 
 ber, when for 83 days the filter-bed and gallery yielded the full 
 (daily) supply to the city of 1,750,000 gallons. In 1878 the filter 
 furnished the full daily supply of 1,879,810 gallons for 43 days. 
 It was found that with a one-inch silt deposit the filter yielded but 
 little water. From August, 1878, to June, 1879, the surface of the 
 bed was not cleaned, and in this time 20 inches of silt had accu- 
 mulated. After being cleaned, it supplied the gallery for only 9 
 days before another inch of silt entirely stopped the yield. Find- 
 ing that the bed needed cleaning three times a month, that the 
 cost would be $25 each time, and that it could only be done when 
 the river was low and free from ice, further cleaning was given up, 
 until 1888, when the new inlet was put in and the bed cleaned, it 
 was practically useless, four-fifths of the water used being pumped 
 directly from the river." 
 
 This filter consisted of a single bed with no provision for peri- 
 odical resting and cleaning. From the description it would seem 
 that the cleaning and restoring of the filter to its original condi- 
 tion were not thought of at the time of its construction. In fact, 
 it was expected to work continuously, and never clog with inter- 
 cepted suspended matter and the products of bacterial action. 
 From the dimensions given, the filter had an " effective area " of 
 0.26 of an acre, and ran for 83 days, delivering water at the rate of 
 6,734,000 gallons per acre per day. At another time it was ope- 
 rated for 43 days at the rate of 7,230,000 gallons per acre per day, 
 when it clogged and was taken out of service. For ten months 
 the filter was run without cleaning, during which time 20 inches 
 of silt accumulated on the bed. 
 
 The statement that it required cleaning three times each month 
 is not remarkable, considering the rate of filtration, but that it 
 cost f 25 to clean one-quarter of an acre of sand filter is remark- 
 able. The cost of cleaning the filters of the East London Water- 
 Works is $25 per acre ; but this is considered there a high price, 
 and is accepted only to secure rapid work under contract, and have 
 the filters out of service for the shortest periods of time. Why 
 it should cost four times as much to clean the Lowell filter is 
 not clear, unless, as it appears, there was no sedimentation of the 
 
TYPES OF SAND FILTERS. 177 
 
 fj 
 
 Merrimac River water before it was put on the bed. Assuming, 
 however, that the filter might have been kept in successful opera- 
 tion at a cost of $ 900 per year, and the average yield of water 
 was 1,800,000 gallons per clay, or 657,000,000 gallons per year, 
 the cost would have been less than 11.40 per million gallons ; a 
 very low price indeed. 
 
 It seems, however, that the filter was not combined with a 
 settling-basin, could only be cleaned " when the river was low and 
 free from ice," and no provision was made for treatment of the 
 river water when this single filter was out of service. 
 
 Considering the time it was built, twenty years ago, and after 
 modern filtration had been established for nearly thirty years in 
 the London Water-Works, it is singular that so many mistakes 
 occurred in the design of this Lowell filter. 
 
 HUDSON, N.Y., FILTERS. 
 
 The following description of these filters is taken from the 
 Manual of American Water-Works, 1889-1890, p. 152. Popula- 
 tion, 10,000. 
 
 Filters. Built 1874-1875. Two of them adjoining the res- 
 ervoir. One with an area of 9,071 square feet, and the other with 
 an area of 23,017 square feet at top of filtering material. Water 
 is admitted to either basin through masonry wells, the walls of 
 which are even with the top of the filtering materials. 
 
 The filtering materials from the top downward consist of : 
 
 6 inches fine white sand. 
 18 " coarse dark sand 
 
 6 " i-inch gravel. 
 
 6 " ^-inch gravel. 
 
 6 " 1-inch broken stone. 
 
 6 " 2-inch broken stone. 
 24 " 4-8-inch broken stone. 
 72 inches total depth of filtering materials. 
 
 "The bottom of the basin has a 6-inch layer of concrete, slop- 
 ing slightly towards center and outlet. A dry masonry stone cul- 
 vert leads along the bottom of basin to an effluent chamber, from 
 
178 THE PURIFICATION OF WATER. 
 
 which water passes to a storage reservoir, or can be drawn directly 
 to the city. When dirty, the top layer of fine sand is removed with 
 flat shovels to a depth of one inch, washed, and replaced. The 
 larger filter was constructed in 1888." 
 
 The average daily consumption of water was stated as 1,483,389 
 gallons, which with an "effective filtering area" of 0.737 acre for 
 the two filters, indicates an average rate of filtration of 2,013,000 
 gallons per acre per day. These filters have been in service since 
 1875. 
 
 POUGHKEEPSIE, N.Y., SETTLING-BASINS AND FILTERS. 
 
 The following description of these filters is taken from the 
 Manual of American Water -Works for 1889-1890, p. 175. Popu- 
 lation, 22,000. 
 
 Settling-Basins and Filters. The former is 30 feet above the 
 mean level of the lower pump-well, 25 by 60 feet, by 12 feet deep, 
 and in three compartments. There are two filter-beds, each 73^ 
 by 200 feet, by 12 feet deep, with 5 feet of filtering materials 
 arranged as follows : 
 
 24 inches sand. 
 
 6 " |-inch gravel. 
 
 6 " 2-inch broken stone. 
 24 " 4-8 inch broken stone. 
 60 inches total depth of filtering materials. 
 
 The materials rest on a concrete floor, in which are open cul- 
 verts conveying the filtered water to an intermediate basin, 6 by 85 
 feet, by 16 feet deep. From this it passes to a reservoir, 28 by 88 
 feet, by 17 feet deep, and from here by 408 feet of 18-inch pipe 
 to the pump-well. 
 
 " The filter-beds cost 54,000. The cost of removing ice, clean- 
 ing beds, and washing sand, in 1888, was $809.00 ; cost per million 
 gallons, $1.32 ; cost of repairs was $86.00." 
 
 Taking the average daily consumption of water at 1,669,358 
 gallons per day, and the effective filtering area as 0.675 acre, the 
 rate of filtration at the date mentioned was 2,473,100 gallons per 
 acre per day. It has doubtless been much higher since 1888-1889. 
 These filters have been in service since 1878. 
 
TYPES OF SAND FILTERS. 179 
 
 
 FILTER GALLERIES. 
 
 These, as exemplified in the water-works of Lyons and Angers, 
 France, are chambers in which the water is collected by infiltration 
 from the surrounding pervious materials. As constructed, they 
 consist of dry masonry walls and covers, placed at an elevation 
 below low water in the adjacent river or other source of supply, 
 and unlike filter beds, the materials through which the water perco- 
 lates to these galleries cannot be conveniently cleaned or graded. 
 Abroad it is believed that these filter galleries collect water from 
 the river or other visible sources ; but experience in this country 
 indicates that such filter galleries, like wells sunk in the bank of a 
 river, generally intercept water percolating through the drift. In 
 fact, such wells sunk in the banks of the Ohio River, within the 
 limestone formation of the channel and watershed, always furnish 
 water quite as hard as that of wells further inland. 
 
 In some situations these galleries doubtless intercept or tap 
 an " underflow " of streams, but such water is not precisely water 
 filtered from these sources, but is the natural percolation of water 
 through the drift parallel to the streams. The remarks in Chapters 
 II. and X. upon natural filtration through the materials of the drift 
 apply to these so-called natural filter galleries. 
 
 From Mr. Kirkwood's Filtration of River Waters * the follow- 
 ing data are taken on the rate of percolation per acre of bottom 
 area of some of these galleries : 
 
 RATE OF PERCOLATION. 
 
 LOCATION. U. S. GALLONS PER ACRE PER DAY. 
 
 Toulouse, France, 12,545,280 
 
 Lyons, 6,403,320 
 
 Angers, 13,068,000 (New gallery.) 
 
 Perth, Scotland, 7,927,920 
 
 The kind of water supplied by these galleries should be that 
 of shallow wells sunk in the drift, and the chemical and bacterial 
 contents may be inferred from the tests of water samples from 
 such wells for any given locality. Aside from the fact that these 
 
 * D. Van Nostrand, N.Y., 1869. 
 
180 THE PURIFICATION OF WATER. 
 
 galleries are liable to intercept surface water insufficiently filtered, 
 and will eventually clog and fail to supply a profitable amount of 
 water, the filtering materials are beyond the reach of daily super- 
 vision and manipulation, and as a general proposition such sources 
 are not to be recommended for domestic water supply. There are 
 localities where filter galleries will furnish an altogether acceptable 
 water supply ; but very careful investigation of the water quality 
 and environment of the source should be made before such water 
 is adopted for drinking and dietetic purposes. 
 
 EUROPEAN FILTERS. 
 
 From the notes collected for the author during 1896, upon some 
 European filters, the following data are taken with reference to 
 the arrangement and nature of the filtering materials, and rates 
 of filtration in U. S. gallons : - 
 
 ROTTERDAM, population, 276,338. 
 
 Water from the River Maas carried for 24 hours in settling- 
 basins before it is put on the filters. 
 
 Area of filtering surface, 9 acres in 18 beds. 
 Standard area of filter, 0.50 acre. 
 
 ARRANGEMENT OF FILTERING MATERIALS. 
 
 Sand at top of bed (effective size, 0.34 mm.),* 30 inches. 
 
 Gravel (over underdrains), 12 " 
 
 Boulders, 12 " 
 
 Total depth of filtering materials, 54 inches. 
 
 Head of water on filters, 42 " 
 
 Average rate of filtration per acre per day, 1,818,200 gallons. 
 
 Average bacteria per c. c. in river water, f 6,000-10,000 
 
 Average bacteria per c. c. in filtrate, 90-99 
 
 Average bacterial reduction by filtration, 98.82 per cent. 
 Average present daily consumption of water, about 7,000,000 gallons. 
 
 Daily per capita consumption, 25 " 
 
 * The effective size of sand-grains in this and the following tables is given on authority of 
 Mr. Hazen. 
 
 t Bacteriological tests made every day. 
 
TYPES OF SAND FILTERS. 181 
 
 
 
 THE .HAGUE, population, 187,545. 
 Water from wells sunk in the sand dunes. 
 Area of filtering surface, 3.66 acres in 6 beds. 
 Standard area of filter, 0.61 acre. 
 
 ARRANGEMENT OF FILTERING MATERIALS. 
 
 Fine dune sand at top of filter (effective size, 0.19 mm.), 30 inches. 
 Gravel, 12 
 
 Boulders, 12 " 
 
 Total depth of filtering materials, 54 inches. 
 
 Head of water on filters, 39 " 
 
 This water is of very excellent quality before it is put on the 
 filter-beds. 
 
 Average rate of filtration per acre per day is 1,497,400 gallons, 
 which is sometimes increased to 4,280,000 gallons per acre per 
 day. 
 
 The thickness of sand-paring is about | inch, the dirty sand 
 being given to truck gardeners for use as a fertilizer. 
 
 Average present daily consumption of water, about 5,480,000 gallons. 
 Daily per capita consumption, 29 " 
 
 AMSTERDAM, population, 489,496. 
 
 Source of water supply : Haarlem sand dunes and River 
 Vecht. 
 
 Area of filtering surface, 5.36 acres in 4 beds, 4.76 acres in 
 7 beds. 
 
 ARRANGEMENT OF FILTERING MATERIALS. 
 
 Fine dune sand at top of bed (effective size, 0.17 mm.), 30 inches. - 
 
 Gravel, 12 
 
 Boulders (over underdrains), 12 " 
 
 Total depth of filtering materials, 54 inches. 
 
 Head of water on filters, 39 " 
 
 Two filters, each having an area of 2,860 square meters (0.706 
 acre), costing, it is stated, $20,000 or 114,200 per acre, have 
 recently been built. 
 
182 THE PURIFICATION OF WATER. 
 
 Rate of filtration per acre per day, 3,208,700 gallons. 
 
 Average daily consumption of water, 10,331,000 " 
 
 Daily per capita consumption, 21 " 
 
 PARIS SUBURBS, population, 600,000. 
 
 Supplied by Compagnie Generate des Eaux, from Choisy-le- 
 
 Roi, Nogent-sur-Marne, and Neuilly-sur-Marne. This company 
 
 uses the Anderson Revolving Iron Purifier and sand filters. 
 
 Area of filters at Choisy-le-Roi, 2.30 acres, 15 beds. 
 
 Area of filters at Neuilly-sur-Marne, 2.30 " 15 " 
 
 Area of filters at Nogent-sur-Marne, 0.75 " 4 " 
 
 Total, 5.35 acres, 34 beds. 
 
 ARRANGEMENT OF FILTERING MATERIALS. 
 
 Fine sand at top of filter, 24 inches. 
 Gravel, 6 " 
 
 Boulders, 6 " 
 
 Total depth of filtering materials, 36 inches. 
 
 Head of water on filters, 36 " 
 
 Rate of filtration per acre per day, I Amwwm 1 g all n s. 
 
 ^ ',OUU,UUU J 
 
 June, 1896, it was claimed that the several stations about Paris 
 where this process was in operation were delivering 17,000,000 
 gallons of filtered water per diem, with a reduction of bacteria 
 from 20,000 per cubic centimeter in the unfiltered water to 
 300-400 in the filtered water (by the Miquel method of cultiva- 
 tion), indicating an efficiency of 98.25 per cent. 
 
 With this iron process the top sand in the filters is worked 
 from a maximum thickness of 24 inches to a minimum thickness 
 of 1*2 inches. The filters are open ; and in winter, when an ice- 
 cake of sufficient thickness is formed, the water level is lowered 
 until the floating sheet of ice rests on the sand, when some of the 
 matter which has accumulated at the surface of the sand and 
 clogged the filter adheres to the under side of the ice, and upon 
 filling the filter again from below and floating the ice, the surface 
 of the sand is opened and the rate of filtration increased. 
 
 A statement is made in connection with the Anderson process 
 at Paris, which, if true, is very significant ; i.e., that the filtered 
 
TYPES OF SAND FILTERS. 183 
 
 
 
 water may be carried in storage for as many as twenty days with- 
 out an increase in the bacterial contents. It is supposed that 
 the salts of iron have an inhibiting effect on the growth of the 
 bacteria, algae, etc., in the treated water. 
 
 ZURICH, population (1892) 93,000. 
 Source of water supply : Lake Zurich and springs. 
 Filtering area, 0.835 acre in 5 beds. 
 
 Three covered filters, 21,690 square feet, 7,230 square feet 
 each. 
 
 Two open filters, 14,967 square feet, 7,483 square feet each. 
 
 ARRANGEMENT OF FILTERING MATERIALS. 
 
 Fine sand at top of bed (effective size, 0.30 mm.), 32 inches. 
 
 Coarse sand, 6 
 
 Small gravel, 4 " 
 
 Coarse gravel (over underdrains), 6 " 
 
 Total depth of filtering materials, 48 inches. 
 
 Head of water on filters, 39.6 " 
 
 Average rate of filtration per acre per day, 5,850,000 gallons. 
 
 Average consumption of filtered water per day, 4,884,000 " 
 
 Average consumption of spring water per day, 792,000 " 
 
 Total average consumption, 5,676,000 gallons. 
 
 Average consumption per capita per day, 61 " 
 
 LONDON. 
 
 According to Mr. Hawksley,* the total area of filters in use 
 by the London companies (1892) was 110 acres, dealing with 
 119,000,000 U. S. gallons per day, from which it appears that the 
 average rate of filtration was about 1,800,000 gallons per acre 
 per day. The average area of the London filters is about one 
 acre each. 
 
 * Appendices to Minutes of Evidence, taken by Royal Commission on Metropolitan 
 Water Supply, 1893, p. 347. 
 
184 THE PURIFICATION OF WATER. 
 
 CHAPTER XII. 
 
 MECHANICAL FILTERS. 
 
 THE mechanical filter is distinctly an American invention, and 
 like many inventions is primarily designed to accomplish a large 
 amount of work within a small compass and short time. 
 
 Certain physical operations have been wonderfully improved in 
 both speed and quality by modern invention, but it cannot be said 
 that this is true of water filtration. While the slowest rates of fil- 
 tration do not invariably give the best results, at the same time 
 certain moderate rates per unit of sand area cannot be increased 
 without risk of impairing the quality of the filtrate. The mechan- 
 ical filter is expected to operate at prodigious rates per unit of sand 
 area, when compared with the very moderate rate of the European 
 type of sand filter. 
 
 This can best be shown by comparison of the rates of the Lon- 
 don filters for 1896 with the estimated best rate for the Morison 
 mechanical filter in the Providence, R.I., tests : 
 
 RATES OF FILTRATION PER ACRE PER DAY. 
 
 U. S. GALLONS. 
 
 Average rate for all the London Filters, 1896, 2,120,000 
 
 Providence Experimental Mechanical Filter, 128,000,000 
 
 The sand used in the mechanical filter is of coarser grain than 
 in the London filters, and this enormous difference of rate could 
 not be maintained were it not for two conditions not found in the 
 operation of the plain sand filter. (1) The frequent washings of 
 the sand-bed, and (2) the use of alum as a coagulant to quickly 
 form on the surface of the sand, the coagulum which takes the 
 place, but cannot be regarded as the equivalent of the " Schmutz- 
 decke," or natural film produced by subsidence of suspended matter 
 and bacterial action at the surface of the plain sand filter. 
 
 The mechanical filter, so-called, is mechanical only so far as 
 
ME CHA NIC A L FIL TERS. 
 
 185 
 
 machinery operated by power is applied for the raking and agita- 
 tion of the sand while being cleansed, and for the regular dosing 
 of the applied water with the alum solution. The process of filtra- 
 tion, excepting as the flocculent alum precipitate may affect it, is 
 altogether natural. 
 
 Mechanical filters may be of two types, those which operate by 
 a gravity head or draft on the sand-bed, and those which operate 
 
 Fig. 11. Jewell Gravity Filter. 
 
 under pressure ; sometimes the full pressure of the water-works 
 system. In either case the machine consists of a tank of wood or 
 metal, vertical or horizontal, in which the bed of sand is carried 
 on a system of screens or strainers, and arranged with a rake or 
 agitator, which is slowly revolved around in the tank through the 
 bed of sand. When the sand is being stirred by the revolving 
 rake a reverse current of water is passed through the filter-bed, to 
 wash away such suspended matter as may have been intercepted 
 
186 
 
 THE PURIFICATION OF WATER. 
 
 Top of Bed -^ 
 
 from the applied water during the previous interval of use of the 
 filter. 
 
 The Gravity Mechanical Filter shown in Fig. 11 is manufac- 
 tured by the O. H. Jewell Filter Company of Chicago, and con- 
 sists of a settling-tank and sand- 
 bed combined in one tank, the 
 lower part containing the settling- 
 tank and the upper part the filter. 
 The head is produced by a vacu- 
 um in the effluent pipes under the 
 sand-bed. In this type of filter 
 the revolving rakes are at rest on 
 the sand-bed while the filter is in 
 service, and are put in rotation 
 after the sand-bed has been "loos- 
 ened up " for washing by a re- 
 verse current of water. Provision 
 is made for washing the sand 
 either with filtered or unfiltered 
 water. 
 
 The most elaborate Report now 
 available on the performance of 
 this type of filter is that made by 
 Mr. Edmund B. Weston, C.E., on 
 what is termed the "Morison Ex- 
 perimental Filter," from which, 
 through the kindness of Mr. J. Her- 
 bert Shedd, city 'engineer of Provi- 
 dence, R.I., the author is permitted 
 to abstract for the purpose of this work. 
 
 The drawing, Fig. 12, is reproduced from the Report men- 
 tioned, in which the filter is described as follows by Mr. Weston : 
 
 " Upon the screens shown at the bottom of the filter, the filtering medium 
 or filter-bed of crushed quartz is located, the total depth being two (2) feet 
 and ten (10) inches. The effective size of the grains of quartz which com- 
 pose the upper two (2) feet is 0.59 mm., and the uniformity co-efficient 1.5. 
 
 Outlet 
 
 Height 14- ft, Diameter 30 inches. 
 
 Fig. 12. Morison Experimental Filter, 
 Providence, R.I. 
 
MECHANICAL FILTERS. 187 
 
 The lower ten (10) inches of quartz is of a much toarser quality. The 
 screens allow the water to pass through them during the different operations 
 of working the filter, downward while filtering, and upward during the pro- 
 cess of washing the filter-bed. They prevent the quartz or any foreign sub- 
 stances from entering the collecting-pipes or passing off with the filtered water. 
 
 " The manner in which the filter was operated during the experiments is 
 as follows : At the end of a run, or immediately before starting the filter, the 
 filter-bed was thoroughly washed by forcing up through the screens and filter- 
 bed a reverse flow of water under pressure, the mechanical rake or agitator, 
 shown in the cut, being operated at the same time, which added materially to 
 the efficient cleansing of the filter-bed. The water was forced up through the 
 bed, and the agitator kept in motion, until the water flowing from the overflow 
 drain-pipe was as clear as it was before it was used for washing the filter-bed. 
 The necessary valves were then operated, and the water and the sulphate of 
 alumina turned onto the filter. 
 
 " The rates of the filtration of water mentioned in this report all represent 
 an average rate per acre per 24 hours unless otherwise specified. The standard 
 rate of filtration decided upon at the commencement of the experiments was 
 128,000,000 gallons per acre per 24 hours. When the term sulphate of alu- 
 mina is used, it is intended as an abbreviation of basic sulphate of alumina. 
 
 "In making the experiments with this filter the following details were 
 carefully investigated, as well as many other points relative to the efficient 
 working of the filter, viz. : 
 
 " 1. The chemicals best adapted for the purification of the Pawtuxet River 
 
 water. 
 " 2. The best method of applying the chemicals, and the quantity to add 
 
 to the applied water for each gallon of water filtered. 
 " 3. If any portion of the chemicals that were added to the applied water 
 
 were present in the filtered water. 
 " 4. The rate in gallons per acre per 24 hours which could be efficiently 
 
 filtered. 
 
 " 5. The bacteriological and chemical purification of the water. 
 " 6. The percentage which the color of the water would be reduced by 
 
 filtration. 
 
 " 7. The washing of the filter-bed. 
 
 " 8. The time which would be required for washing the filter-bed. 
 " 9. The quantity of water which would be required to wash the filter-bed. 
 " 10. The quantity of water which it would be necessary to run to waste 
 
 after washing the filter-bed. 
 
 " 11. The length of time which the filter would run after starting, before it 
 would be necessary to shut down and wash the filter-bed on ac- 
 count of the water gradually rising to its prescribed limit in the 
 filter, owing to the filter-bed becoming gradually clogged up. 
 
188 THE PURIFICATION OF WATER. 
 
 " 12. The effective stability of the quartz and supplementary precipitate 
 bed; viz., whether it could be depended upon to do its work 
 thoroughly during the whole of the time that the filter was in 
 operation, or whether at times it would be liable to crack or 
 break, or have its efficiency reduced in any manner. 
 
 "13. The loss of head due to the water flowing through the filter-beds 
 and screens. 
 
 " During the preliminary experiments, the chemicals used were basic sul- 
 phate of alumina, chloride of alumina, carbonate of soda, bicarbonate of soda, 
 caustic soda, and chloride of iron. The soda salts were used in connection 
 with sulphate of alumina. It was found, however, that basic sulphate of alu- 
 mina added to the applied water produced the best results. Basic sulphate of 
 alumina, therefore, is the only chemical that has been used since the prelimi- 
 nary experiments. 
 
 " The theory of mechanical filtration, when basic sulphate of alumina has 
 been added to the applied water, may be described as follows : The alumina 
 causes an artificial precipitation. A portion of the alumina is decomposed, 
 forming sulphates of other bases and a flocculent precipitate of aluminic 
 hydrate. A portion of it also combines directly with the organic matter 
 present in the water, coagulating the same, and thus helping to increase the 
 precipitation. A sufficient quantity of the precipitate having been deposited 
 upon the top of the sand or quartz-bed of the filter, and plugged into the in- 
 terstices of the upper layer of sand or quartz-grains, the filter is ready for 
 service. 
 
 " At the commencement of the experiments with the Morison mechanical 
 filter, it was discovered that satisfactory results could not be obtained by sim- 
 ply dropping the sulphate of alumina into the applied water at the rate of \ 
 grain per gallon, as it would take from one to three hours after the filter was 
 started for a sufficient quantity of the precipitate to form in order to do good 
 work. After experimenting in different ways, it was found that if a "free 
 flow " of about a pint of coagulant, containing about nine hundred and eleven 
 (911) grains of sulphate of alumina for an average rate of filtration of about 
 128,000,000 gallons per acre per 24 hours, was allowed to run into the filter, 
 immediately after the water was let on, in a space of time of not more than 
 six (6) minutes, a quantity of coagulant corresponding to one-half () grain of 
 sulphate of alumina per gallon of filtered water being dropped in at the same 
 time from a different receptacle than that containing the "free flow," a suffi- 
 cient amount of precipitate would be formed to do good work in one-half hour 
 or less after the water commenced to flow from the filter. 
 
 " At the commencement of a run of the filter, the applied water was at 
 first gradually let into the filter, it being regulated at the same time. After 
 the normal quantity commenced to flow into the filter a constant flow was 
 maintained, and the depth of water in the filter gradually increased proportion- 
 
MECHANICAL FILTERS. 189 
 
 ately during the run as the supplementary precipitate bed. was formed, and 
 the filter-bed became plugged with precipitate. The rise of water practically 
 accommodated itself to the circumstances, and caused a constant flow of 
 water through the filter, which I considered extremely essential in order to 
 obtain good results. 
 
 " One of the most serious problems that it was necessary to solve when 
 the experiments were commenced, was to ascertain if the basic sulphate of 
 alumina that was added to the applied water was entirely decomposed before 
 the water was discharged from the filter." 
 
 A sample of the sulphate of alumina used had the following 
 composition : - 
 
 ONE-HALF () GRAIN 
 PER CENT. 
 
 CONTAINS IN GRAINS. 
 
 Insoluble residue, 0.52 0.0026 
 
 Alumina (A1 2 O 8 ), 15.78 0.0789 
 
 Sulphur Trioxide (SO 3 ), 36.79 0.1840 
 
 Water (by difference), 46.91 0.2345 
 
 100.00 0.5000 
 
 After a series of experiments upon this filter, extending from 
 April 5, 1893, to Jan. 80, 1894, Mr. Weston reached the follow- 
 ing conclusions : 
 
 The chemical best adapted for the purification of the Pawtuxet 
 River water was basic sulphate of alumina, the quality used con- 
 taining 15.8 to 17.5 per cent of alumina. 
 
 The best method of applying the chemicals, and the quantity 
 required per gallon of water, have been described as \ grain per 
 gallon plus "free flow " for not more than six minutes, equivalent 
 to T % grain of sulphate of alumina per gallon of water for an aver- 
 age run of 16 hours, 43 minutes. 
 
 Upon the appearance of any portion of the chemicals added to 
 the water in the filtrate, he says : 
 
 " The results that I have mentioned, that were obtained by applying the 
 logwood and acetic acid test for alum, in conjunction with filter paper, have 
 demonstrated, I think, that none of the basic sulphate of alumina was present 
 during the experiments in the filtered water, in rts original state, after the water 
 had been flowing from the filter twenty-one (21) minutes. The only indica- 
 tion of alumina found in the filtered water was a minute quantity of finely sus- 
 pended hydrate, resulting from the addition of the alumina, that came through 
 the filter-bed with the water that was being filtered. . . . 
 
190 THE PURIFICATION OF WATER. 
 
 " An analysis by Professor Thomas M. Drown . . . shows that 0.0292 of a 
 grain of alumina (A1 2 O 3 ) per gallon was found in a sample of Pawtuxet River 
 water, that had been taken directly from the river, and afterwards filtered 
 through a double thickness of Swedish paper, and that 0.0584 of a grain of 
 alumina (A1 O 3 ) per gallon was found in a sample of the same water after 
 sulphate of alumina had been added to it, at the rate of one-half (A) grain per 
 gallon ; and the very slight flocculent precipitate produced, filtered off through a 
 double thickness of filter paper, shows an increase of alumina (A1 2 O 3 ) of 
 0.0292 of a grain." 
 
 The rate at which the water was " filtered successfully " ranged 
 from 90,000,000 gallons to 193,000,000 gallons per acre per 24 
 hours, "the average rate of nitration being about 128,000,000" 
 gallons per acre. 
 
 The average bacterial efficiency of the filter for two short in- 
 tervals of time, selected from the ten months of experiment, viz., 
 Oct. 17 to Nov. 11, 1893, a period of 25 days, and from Jan. 24 to 
 Jan. 30, 1894, a period of 7 days, was a reduction of 98.6 per cent 
 of the bacterial contents of the applied water. For various rea- 
 sons Mr. Weston rejects the work of the filter for 87 per cent of 
 the time of test, and bases the bacterial efficiency upon these two 
 short intervals of time. Upon the subject of bacterial efficiency, 
 Mr. Weston says : 
 
 " I do not consider that the efficiency of a filter should be entirely based 
 upon the average results obtained, although this is generally the standard upon 
 which the efficiency is based, but that the worst results obtained should be 
 duly considered. In order to present my ideas upon this subject more clearly 
 I will assume a rather improbable case. For example, if 100 individual results 
 were used in working up an average, 90 of these results might each show an 
 efficiency of 100 per cent, and 10 of them might each show an efficiency of 
 only 80 per cent, or in other words, 10 per cent of the total results would be 18 
 per cent below the average result, which in my opinion would be sufficient 
 grounds to condemn a filter. Yet the average of the whole number would be 
 98.0 per cent, which is a very good result." 
 
 The author concurs with Mr. Weston, with the additional sug- 
 gestion that the efficiency of a filter for hygienic purposes should 
 be measured altogether by its worst results, and not by the best 
 or even the average results. 
 
 The filtration of a public water supply should assume that under 
 
MECHANICAL FILTERS. 191 
 
 the most unfavorable conditions of the applied water, the filtrate 
 shall comply with a given standard of hygiene, and any estimate 
 of the influence of such water upon the health of the people who 
 drink it should be based upon these worst conditions. In short, 
 the nitration of a public water supply should assume a certain 
 uniformity in the quality of the nitrate, without regard to the man- 
 ual operation of the filters, or to the condition of applied water. 
 The filter itself should be so constructed, and the regulations under 
 which it is operated such, that the quality of the filtrate shall sat- 
 isfy some acceptable standard at all times. Phenomena and phys- 
 ical aberrations may be tolerated in scientific investigations, but 
 the practical purification of a polluted public water supply should 
 involve no phenomena and no vagaries to the prejudice of the 
 public health. 
 
 The reduction of color by filtration, using alum with the " Mor- 
 ison filter," ranged from 66.3 per cent for the night observations 
 to 77.9 per cent as an average for the day observations. 
 
 The average time required to wash the bed of sand in the filter 
 was about 11 minutes ; and the amount of water required to wash 
 the sand-bed, and the amount which was run to waste after the 
 filter was washed and started in service, was about 7.8 per cent of 
 the quantity filtered. Of this quantity 4.9 per cent was required 
 to wash the filter-bed, and 2.9 per cent was run to waste after the 
 filter was started again. 
 
 The average length of "run" of the filter between cleansings 
 was 16 hours and 43 minutes, and the average loss of head 
 for a delivery of 128,000,000 gallons per acre per diem of 24 
 hours was 2.44 feet. Mr. Weston estimates the cost of ope- 
 rating a Morison mechanical filter plant of 15,000,000 U. S. gal- 
 lons daily capacity, as $5.69 per one million gallons of water 
 filtered. 
 
 From a single experiment by Dr. T. M. Drown, in connection 
 with the Providence filter tests, it appears that the addition of 
 one-half () grain per gallon of the sulphate of alumina to the 
 Pawtuxet River water, increased the alumina from 0.0292 grain 
 per gallon to 0.0584 grain per gallon, and increased the sulphuric 
 acid in the water from 0.3129 to 0.5214 grain per gallon. 
 
192 THE PURIFICATION OF WATER. 
 
 Although Mr. Western has made a very long and exhaustive 
 study of the mechanical filter, with "alum " as a coagulant, in his 
 summary he rejects the majority of his data, and draws his con- 
 clusions from : 
 
 Two days' operation of the filter in July. 
 Seven days' operation of the filter in October. 
 Eight days' operation of the filter in November. 
 Two days' operation of the filter in December. 
 Twenty-four days' operation of the filter in January. 
 
 Or out of ten months' continuous experiment, he takes 43 days, 
 as showing the possibilities of this method of water filtration ; of 
 which time, six days embrace a treatment of the sand-bed hitherto 
 untried with any type of filter. The average amount of alum 
 used during these 43 days was about T 7 o grain per gallon of 
 water. 
 
 So far as the chemical and bacterial reductions in the applied 
 water are concerned, it may be accepted that the mechanical filter, 
 used with " alum " as a coagulant, will accomplish about the same 
 results as will natural filtration ; but with the latter no injurious 
 property can be imparted to the water, while with "alum " proper- 
 ties more objectionable than the pollution sought to be removed 
 may be found in the filtrate. 
 
 The following costs of construction and operation of mechanical 
 and plain sand filters are taken from estimates made by Mr. 
 Weston for the city of Providence, R.I. The cost of filters is 
 based on an available daily capacity of 15,000,000 gallons. 
 
 MECHANICAL FILTERS. 
 
 NUMBER AND KIND. 
 
 COST OF FILTERS 
 PER MILLION GALLONS OF 
 DAILY CAPACITY. 
 
 COST OF FILTRATION 
 PER MILLION GALLONS OF 
 WATER TREATED. 
 
 60 Steel Filters, 
 60 Cypress " 
 51 Steel 
 51 Cypress " 
 
 $16,344.80 
 15,296.80 
 14,160.30 
 13 262.30 
 
 $7.67 
 7.86 
 7.25 
 7.41 
 
MECHANICAL FILTERS, 
 
 193 
 
 PLAIN SAND FILTERS. 
 
 KIND OF FILTERS. 
 
 COST OF FILTERS 
 PER MILLION GALLONS OF 
 DAILY CAPACITY. 
 
 COST OF FILTRATION 
 PER MILLION GALLONS OF 
 WATER TREATED. 
 
 Filters with vaulted 
 masonry coverings, 
 Filters with timber coverings, 
 Open filters, 
 
 $35,000.00 
 21,906.53 
 19,414.66 
 
 $8.86 
 8.14 
 6.87 
 
 In the figures for cost of operation of the mechanical filters, 
 the cost, including alum, is taken in all cases at 4.52 per million 
 gallons of water treated ; and the cost of operation of the plain 
 sand filters, as deduced by Mr. Weston from the Reports of the 
 Massachusetts State Board of Health at Lawrence, is taken at 
 84.39 per million gallons of water treated. 
 
 The cost per million gallons of water filtered is taken in the 
 table as the whole cost, including interest at four per cent on cost 
 of filters, buildings, and all appurtenances, and charges for deterio- 
 ration of plant, cost of chemicals (for mechanical filters), and all 
 labor required to operate the works. 
 
 Other uses of the mechanical filter, in several cities in New 
 Jersey, are given in an interesting paper by Mr. M. N. Baker, C.E., 
 contributed to the New Jersey Sanitary Association, 1895, from 
 which the following quotations are made : 
 
 "Without attempting to trace the various stages through which the me- 
 chanical filter has passed, it may be said that it aims to purify large volumes 
 of water with a small body of filtering material, relying upon frequent wash- 
 ings to keep the material clean." 
 
 " Mechanical filters are now used on over one hundred American water 
 supplies, against perhaps ten filter-bed plants worthy the name. Many factors 
 have contributed to the greater use of mechanical than slow sand filters in this 
 country. Chief of these is the commercial aspect of mechanical filtration and 
 a former entire misconception of the principles of, and the results which may 
 be accomplished with, filter-beds. Mechanical filters have been vigorously 
 pushed by sales agents wherever bad water has been reported. Filter-beds 
 stand or fall on their own merits, as they are not patented, and no one is finan- 
 cially interested in securing their adoption. The misconceptions regarding 
 filter-beds have come to light with the recognition of the germ theory of dis- 
 ease, and improved methods and interpretation of the bacterial examinations 
 
194 THE PURIFICATION OF WATER. 
 
 of water. For years water contained disease and other germs, and filter-beds 
 removed many of them without any one knowing it. 
 
 " Most of the plants built by the mechanical filter companies have been 
 designed to remove suspended matter and color where these were so marked 
 as to render the water almost intolerable. These ends many of the mechan- 
 ical filters have accomplished most admirably. Meanwhile, most communities 
 have remained satisfied with water that looks well, without regard to the dan- 
 gerous impurities it may contain, or have secured new supplies from more 
 favorable sources. With the modern advances in sanitary science, attention 
 has recently been turned to the importance of removing sewage impurities from 
 polluted waters, if the latter must be used ; and of late filter-beds and mechan- 
 ical filters have been constructed with this end in view. 
 
 " I believe it probable that good bacteriological results are possible with 
 mechanical filtration. I am certain that they can be obtained with sand filter- 
 beds, and that suspended matter, vegetable stains, and iron can be removed 
 by means of mechanical filters. 
 
 " Filter-beds will also remove color, suspended matter, and iron, if supple- 
 mented by aeration, as well as sewage impurities." 
 
 The following data from the Somerville and Raritan Water- 
 Works are given by Mr. Baker: 
 
 Source of supply, Raritan River. 
 
 Population supplied (1890), 6,417. 
 
 Capacity of filter, 1,500,000 gallons per day. 
 
 Average daily filtration for 1894, about 800,000 gallons. 
 
 Cost of filters, exclusive of buildings, $15,600. Of the opera- 
 tion of these filters, Mr. Baker says : 
 
 " As this plant was put in to remove matters in suspension, its efficiency 
 should be judged by its removal of total solids. These were reduced from 
 26.72 to 15.98 parts per 100,000, a reduction of 10.74 parts in 26.72, or about 
 42 per cent. The color is reported as having been changed from dark brown 
 to faint. The reduction of organic matter, as indicated by the albuminoid am- 
 monia, was nearly 70 per cent." 
 
 The analyses from which these figures are taken contain meas- 
 urable quantities of nitrites and nitrates for the river water, while 
 none are given for the 'filtered water. Successful filtration gene- 
 rally indicates an increase of nitrites and nitrates. Referring to 
 the ammonias, the albuminoid ammonia is reduced from .049 to 
 .015, and the " free " ammonia is increased from .013 to .052 part 
 per 100,000. 
 
MECHANICAL FILTERS. 
 
 195 
 
 The permanent hardness of the water by addition of sulphates 
 was increased from 3.75 parts to 6.75 parts per 100,000, while the 
 temporary hardness (carbonates) was reduced from 3.50 parts to 
 0.50 part per 100,000, the gain in sulphates being exactly bal- 
 anced by the loss in carbonates. Alum is used as the coagulant. 
 
 Long Branch, according to Mr. Baker, has a mechanical filter 
 plant of 3,000,000 gallons daily capacity, treating water from Whale 
 Pond Brook. These are pressure filters, working under a head of 
 40 pounds. The loss of head in passing the water through the 
 filter is stated at five-tenths pound. Five per cent of the total 
 water purified is required for washing the sand in the filters. The 
 filters, including buildings, cost $31,000. 
 
 From the analyses given in Mr. Baker's paper the following 
 data are taken : 
 
 PARTS PER 100,000 OF WATER. 
 
 BEFORE FILTRATION. 
 
 AFTER FILTRATION. 
 
 Free ammonia, 
 
 0.132 
 
 0.0035 
 
 Albuminoid ammonia, 
 
 0.0445 
 
 0.0095 
 
 Nitrites, 
 
 0.0025 
 
 0.0015 
 
 Nitrates, 
 
 0.087 
 
 0.087 
 
 Permanent hardness, 
 
 . . . 
 
 1.00 
 
 Temporary " 
 
 2.25 
 
 1.24 
 
 Total solids, 
 
 9.52 
 
 . 7.14 
 
 Organic and volatile matters, 
 
 4.24 
 
 1.74 
 
 Bacteria per c. c., 
 
 268.00 
 
 3.0 
 
 Potash alum is used in this filter. 
 
 THE JEWELL GRAVITY FILTERS AT LORAIN, OHIO. 
 
 Through the kindness of the Jewell Filter Company of Chicago, 
 the author is enabled to include the results of its process of water 
 purification at Lorain, Ohio, for the months of March and April, 
 1897. From the notes furnished, which are quite complete, the 
 following resumt is drawn : 
 
 MARCH, 1897. 
 
 Bacteria in untreated water, 3,725 
 
 " " filtered water, _ 88 
 
 Percentage of reduction, 97.61 
 
 Grains of alum used per gallon of water, 1.87 
 
 Average water treated daily, 2,712,400 gallons. 
 
196 THE PURIFICATION OF WATER. 
 
 APRIL, 1807. 
 
 Bacteria in untreated water, 1,835 
 
 " " filtered water, 31 
 
 Percentage of reduction, 98.29 
 
 Grains of alum used per gallon of water, 1.99 
 
 Average water treated daily, 2,774,500 gallons. 
 
 In this case the alum per gallon of water treated is about 
 2 grains, which, at IT G O cent per pound, represents a cost of |4.80 
 per million gallons of water for the chemical alone, an amount 
 equal to the average cost of plain sand nitration, according to the 
 reports from works where such filters are in daily use. The actual 
 consumption of alum per gallon of water is over three times the 
 amount which Mr. Weston, in the report on mechanical niters 
 for the city of Providence, regarded as necessary for proper fil- 
 tration. 
 
 Adding to the cost for alum the estimate by Mr. Weston 
 of the cost for labor, wash-water, water run to waste, etc., viz., 
 $2.80, the total cost of mechanical nitration, based on the Lo- 
 rain experience, will be $7.60 per one million gallons of water 
 treated. 
 
 According to Mr. James H. Blessing, of the New York Filter 
 Company, in a statement to the city of Albany, N.Y.,* the cost 
 for mechanical filtration at that place, including labor, alum, etc., 
 will be from $2.44 to $2.94 per million gallons. This estimate is 
 very difficult to reconcile with the estimated cost at Providence, 
 and the probable actual cost at Lorain ; in fact, Mr. Blessing 
 makes the cost for Albany from one-half to two-thirds the cost 
 for alum alone at Lorain, and considerably less than one-half of 
 the total cost as figured by Mr. Weston for the proposed me- 
 chanical filter plant for Providence, R.I. 
 
 MECHANICAL FILTERS, ALBANY, N.Y. 
 
 Mr. Hazen,f in his report on filtration for Albany, estimates 
 the cost of mechanical filtration as follows : 
 
 * An Address to the Common Council, Feb. 27, 1897. 
 
 t Report to Board of Water Commissioners, Albany, Feb. 13, 1897, p. 27. 
 
MECHANICAL FILTERS. 
 
 197 
 
 COST OF FILTRATION PER MILLION GALLONS. 
 
 LOCATION OF FILTERS. 
 
 LUMBER DISTRICT. 
 
 BLEECKER RESERVOIR. 
 
 Labor and power for operation of filters, 
 Alum, 143 Ibs. at 1.6/, 
 Wash-water, 
 
 $1.50 
 2.29 
 
 $1.50 
 2.29 
 0.50 
 
 Total, 
 
 3.79 
 
 4.29 
 
 Mr. Blessing's estimates, 
 
 2.44 
 
 2.94 
 
 The estimated cost of mechanical filters for Albany is given as 
 follows : 
 
 COST OF MECHANICAL FILTERS PER MILLION GALLONS OF DAILY CAPACITY. 
 
 LOCATION. 
 
 HAZEN. 
 
 BLESSING. 
 
 Lumber district, 
 Bleecker Reservoir, 
 
 $12.938.24 
 12,118.22 
 
 $ 6,064.42 
 10,431.37 
 
 The cost of filtration quoted for Albany does not include (as 
 in the report) the cost of lifting the water to the filters, which will 
 vary in different locations, and with gravity sources may not be 
 required at all. The cost of filters includes the pro rata allowance 
 for contingencies. Mr. Hazen estimates on gravity filters for both 
 locations. Mr. Blessing estimates on gravity filters for the lum- 
 ber district, and on pressure filters for the Bleecker reservoir. 
 
 The prices for filters per million gallons of daily capacity do 
 not include land, nor such structures as are required at Albany to 
 make the filtered water available in service. 
 
 MECHANICAL FILTRATION FOR PHILADELPHIA. 
 
 From a proposition of the Morison-Jewell Filtration Company 
 to the city of Philadelphia * the following data is by permission 
 extracted : 
 
 Daily capacity of filters, 30,000,000 gals. 
 
 Rate of filtration per acre of sand area per day, 128,000,000 " 
 Time allowed for subsidence of water in settling-tanks, One hour. 
 
 Total cost of filters, foundations, and buildings, $300,000.00 
 Cost per 1,000,000 gallons of capacity, $10,000.00 
 
 * June 8, 1897. 
 
 OF THB 
 
 UNIVERSITY 
 
198 THE PURIFICATION OF WATER. 
 
 The manufacturers' estimated cost for operation of these filters, 
 with an allowance of f grain of sulphate of alumina per gallon of 
 water, is given as $3.61 per one million gallons of water filtered, 
 which would be increased to $6.43 per million gallons if the con- 
 sumption of alum should be as high as 2 grains per gallon. 
 
 MECHANICAL FILTERS, ELMIRA, N.Y. 
 
 The city of Elmira, N.Y., is supplied with water from Morison- 
 Jewell mechanical filters, using sulphate of alumina as a coagulant. 
 These filters have a daily capacity of 6,000,000 gallons, and are 
 operated at the rate of 100,000,000 gallons per acre per day of 24 
 hours. The water passes through the settling-tanks by continuous 
 flow at a rate which is equivalent to the detention of the raw river 
 water in the tanks for about 30 minutes before it passes to the 
 filters. Dr. Ravenel,* of the University of Pennsylvania, has made 
 the following bacteriological analyses of the water before and after 
 filtration. 
 
 The first of the following tables gives the bacterial results for 
 mechanical filtration without a coagulant, while the second shows 
 the bacterial efficiency with basic sulphate of alumina used at the 
 rate of 1.4 grain per gallon of water : 
 
 BACTERIAL CONTENTS OF WATER, WITHOUT COAGULANT. 
 
 DATE. 
 
 RAW WATER. 
 
 AFTER PASSING 
 SETTLING-TANK. 
 
 FILTERED WATER. 
 
 June 10, 1897. 
 
 885 per c. c. 
 
 746 per c. c. 
 
 36 per c. c. 
 
 11 " 
 
 513 
 
 556 
 
 Samples lost. 
 
 12 " 
 
 625 
 
 527 " 
 
 40 per c. c. 
 
 13 
 
 351 " " 
 
 208 " 
 
 132 " 
 
 14 
 
 434 
 
 366 " " 
 
 34 " 
 
 15 
 
 425 
 
 216 
 
 13 " 
 
 16 
 
 566 " " 
 
 407 " " 
 
 15 " 
 
 17 " 
 
 432 " 
 
 358 " " 
 
 21 " 
 
 BACTERIAL CONTENTS OF WATER, USING BASIC SULPHATE OF ALUMINA 
 1& GRAIN PER GALLON OF WATER. 
 
 DATE. RAW WATER. sSuN. FILTERED WATER. 
 
 June 10, 1897. 885 per c. c. 576 per c. c. 26 per c. c. 
 
 11 513 " " 413 " 40 " 
 
 12 625 457 " " 12 " 
 Communicated by the Morison- Jewell Filtration Company, New York, July, 1897. 
 
MECHANICAL FILTERS. 199 
 
 DATE. 
 
 RAW WATER. 
 
 AFTER MASSING 
 SETTLING-TANK. 
 
 FILTERED WATER. 
 
 June 13, 1897. 
 
 351 per c. c. 
 
 129 per c. c. 
 
 10 per c. c. 
 
 14 " 
 
 434 
 
 258 
 
 " " 
 
 15 
 
 425 " 
 
 333 
 
 6 " 
 
 16 
 
 566 " 
 
 315 
 
 13 
 
 17 
 
 432 
 
 117 
 
 12 " 
 
 In these tests the water was taken from the Chemung River, 
 and carried for short intervals of time in settling-tanks. For the 
 first test without, and for the second test with a coagulant. 
 
 The average numbers of bacteria, and bacterial efficiencies 
 of sedimentation and filtration, without a coagulant, were as fol- 
 lows : 
 
 Average bacteria in river water, 529 per c. c. 
 
 Average bacteria in subsided water, 423 " " 
 
 Average bacteria in filtered water, 26.5 " " 
 Reduction of bacteria by subsidence, 20 per cent. 
 
 Reduction of bacteria by subsidence and filtration, 95 " " 
 
 (In striking the average of bacteria per c. c. for the filtered 
 water without a coagulant, the determination for June 13 is 
 omitted ; as the figures given clearly indicate an abnormal result, 
 the cause of which is not explained in the excerpt of the report 
 in possession of the author.) 
 
 In the second test a coagulant was used, with the following 
 average numbers of bacteria per c. c. of water and bacterial 
 efficiencies, by subsidence and filtration : 
 
 Average bacteria in river water, 529 per c. c. 
 
 Average bacteria in subsided water, 325 " " 
 
 Average bacteria in filtered water, 15 " " 
 
 Reduction of bacteria by subsidence, 38.6 per cent. 
 
 Reduction of bacteria by subsidence and filtration, 97.16 " 
 
 The addition of sulphate of alumina to the Chemung water 
 increased the efficiency of subsidence and filtration 2.16 per cent, 
 but in this instance the percentage reductions are of minor 
 importance to the low numbers of bacteria found in the fil- 
 trates. 
 
200 THE PURIFICATION OF WATER. 
 
 THE USE OF ALUM FOR FILTRATION. 
 
 If the decomposition of alum in mechanical filters depends 
 upon the amount of bases, as lime, soda, etc., in the water, then 
 no free or unappropriated sulphuric acid can at any time be pres- 
 ent, because all such will be found in combination with the alu- 
 mina (alum), or with the lime, etc., as sulphates ; and in this 
 case an excess of alum applied to the water would result in a 
 hydrated sulphate of alumina, some of which may appear in the 
 nitrate. Alum can be dissolved in distilled water wholly free from 
 organic matter or earthy salts ; but no decomposition of the alum 
 will in such case occur, because of the lack of a base to appro- 
 priate or unite with the sulphuric acid. Astringency would be 
 imparted to the water ; and a physiological question then arises, 
 upon the effect on the absorbent vessels of the digestive tract, of 
 the continuous use of water containing an astringent. 
 
 So far as information from medical sources has come to the 
 author upon this question, it indicates an objection to the continu- 
 ous use of a drinking-water purified by alum ; the disorders trace- 
 able to it being impaired digestion, irritation of the mucous 
 membrane of the stomach, and when gastric troubles already exist, 
 a dangerous aggravation of these may follow the continuous use 
 of water containing perceptible astringent properties. For the 
 bath and laundry, and for some industrial purposes, water puri- 
 fied by an addition of alum is well known to be objectionable. 
 
 When mechanical filters are used for the treatment with alum 
 of polluted soft waters, as are the waters of many of the rivers 
 of the central and western portions of the United States, lime as 
 milk of lime is sometimes added to the water before the alum is 
 introduced, to furnish a base for the sulphuric acid in the sulphate 
 of alumina to unite with. This practice, the author is informed, is 
 in use in several water- works supplied with Jewell filters, with an 
 improvement over the use of alum alone, in the quality of the 
 filtrate, and a reduced cost for chemicals. In the case of one city 
 using Mississippi River water, it is reported that, by the addition 
 of lime to the water, the work of the filters is more regular, and 
 the consumption of sulphate of alumina kept within 2 grains per 
 million gallons of water treated. 
 
MECHANICAL FILTERS, 201 
 
 
 
 The superintendent of a Western water- works, in writing to 
 the author, says : 
 
 " When the river is soft, which is the case after heavy rains, we use lime- 
 water in the pump-well, thus supplying carbonate of lime for the alumina to 
 act upon. This has proven an economical measure, as less alumina is required 
 when the carbonates are present in large quantities, and compared with sul- 
 phate of alumina, lime is inexpensive." 
 
 The published circular of one of the prominent manufacturers 
 of mechanical niters contains the statement : " As a general rule, 
 when operating niters at full capacity . . . the amount of alum 
 . . . required varies from T V to 2 grains per gallon," and "within 
 certain limits the alum required is inversely proportional to the 
 rate of filtration." The smaller amount of alum mentioned (yV 
 grain per gallon) is very much less than the amount reported to 
 the author from any public water-works which employs mechan- 
 ical nitration. Indeed, the larger amount (2 grains per gallon) 
 seems to more nearly represent the consumption of alum in 
 practice. 
 
 The author has been informed by manufacturers and users 
 of mechanical filters, that one of the difficulties with the use of 
 alum in the waters of our Western rivers is the variable quantity 
 required for good results from day to day. That while \ grain 
 or less per gallon of water would be sufficient one day, 6 grains 
 would be necessary to obtain satisfactory results another day. 
 This represents not only a large cost for alum, but raises a ques- 
 tion of the reliability of a process of water purification subject to 
 such a wide range of behavior in actual service ; and this large 
 variation in the quantity of coagulant required, it is stated, is 
 not always accompanied by known corresponding changes in the 
 quality of the unfiltered water. 
 
 The table on page 202 shows the influence of rate of filtration 
 and variable quantities of alum per gallon of applied water on the 
 bacterial efficiency of the Lorain filters.* 
 
 Before dismissing this subject, the author desires to quote a 
 pertinent paragraph in conclusion of a report to the Philadelphia 
 
 * Ohio Sanitary Bulletin, Columbus, Ohio, October, 1897, p. 117. 
 
202 
 
 THE PURIFICATION OF WATER. 
 
 LORAIN, OHIO, MECHANICAL FILTERS. 
 (From Examinations by MR. F. S. HOLLIS.) 
 
 
 
 BACTERIA PER C.C. OF WATER. 
 
 PERCENTAGES. 
 
 RATE OF FILTRATION. 
 
 ALUM. 
 
 
 
 
 BACTERIA 
 
 GALLONS PER ACRE 
 PER DAY. 
 
 GRAINS 
 PER GALLON. 
 
 LAKE WATER. 
 
 FILTRATE. 
 
 BACTERIAL 
 REDUCTION. 
 
 REMAINING 
 
 IN 
 
 
 
 
 
 
 FILTRATE. 
 
 66,489,984 
 
 2.58 
 
 1,441 
 
 16 
 
 98.9 
 
 1.1 
 
 68,999,040 
 
 2.50 
 
 385 
 
 6 
 
 98.4 
 
 1.6 
 
 69,626,304 
 
 2.27 
 
 367 
 
 9 
 
 97.5 
 
 2.5 
 
 80,289,792 
 
 1.07 
 
 154 
 
 14 
 
 90.9 
 
 9.1 
 
 71,508,096 
 
 0.94 
 
 189 
 
 26 
 
 86.3 
 
 13.7 
 
 Water Department, by Drs. N. Wiley Thomas and John Marshall, 
 on the subject of "alum " filtration for the city water supply : 
 
 " It appears practically impossible to rapidly filter the city's supply with- 
 out the use of a coagulant ; and while any method of filtration must of neces- 
 sity be largely experimental, yet in view of the unsatisfactory results of our 
 examination of the water obtained from the Roeske filter, and in consideration 
 of the approximation of the Long Branch filter to the spirit of the specifica- 
 tions the water after treatment being sensibly improved (although it does 
 not literally fulfill the conditions in the specifications stated, yet it might pos- 
 sibly do so at Belmont), we beg leave to suggest the erection and operation of 
 a plant of the character proposed by the New York Filter Company at the 
 Belmont Water-Works, under the considerations proposed, provided that every 
 possible precaution be taken to prevent an excessive use of alum, if this sub- 
 stance be employed as a coagulant ; and, if the sulphate of alumina be selected, 
 that corresponding care be exercised that it shall be free from dangerous im- 
 purities, and shall be introduced only in sufficient amount to produce the 
 necessary coagulation ; and finally that limestone be made a part of the filter- 
 bed, to insure the presence of an adequate amount of lime compounds not 
 already converted into sulphate, to take up the products of the decomposition 
 of the alum, as well as to facilitate the breaking up of that compound." 
 
 An examination of the Manual of American Water- Works for 
 1897 reveals the use of various kinds of filters in the water-works 
 of one hundred and sixty-one cities and villages. About two- 
 thirds of this number are represented by the different forms of 
 mechanical filters, of which the largest works are collected in 
 the table on page 203. The daily aggregate capacity of all the 
 mechanical filters in water-works of the United States is about 
 190,000,000 gallons. 
 
MECHANICAL FILTERS. 
 
 203 
 
 CITY. 
 
 DAILY 
 CAPACITY IN 
 
 GALLONS. 
 
 CITY. 
 
 DAILY 
 CAPACITY IN 
 GALLONS. 
 
 Wilkes-Barre, Pa., 
 
 10,000,000 
 
 Cedar Rapids, Iowa, 
 
 4,000,000 
 
 Chattanooga, Term., 
 
 9,000,000 
 
 Elgin, III, 
 
 4,000,000 
 
 Davenport, Iowa, 
 
 7,500,000 
 
 Newport, R. I., 
 
 4,000,000 
 
 Atlanta, Ga., 
 
 7,000,000 
 
 Burlington, Iow r a, 
 
 3,500,000 
 
 Elmira, N. Y., 
 
 6,000,000 
 
 Biddeford and Saco, Me., 
 
 3,000,000 
 
 Oakland, Cal., 
 
 6,000,000 
 
 Columbia, S. C., 
 
 3,000,000 
 
 Quincy, 111., 
 
 5,000,000 
 
 Decatur, III, 
 
 3,000,000 
 
 Bangor, Me., 
 
 5,000,000 
 
 Greenwich, Conn., 
 
 3,000,000 
 
 Little Rock, Ark., 
 
 5,000,000 
 
 Keokuk, Iowa, 
 
 3,000,000 
 
 Knoxville, Tenn., 
 
 5,000,000 
 
 Lorain, Ohio, 
 
 3,000,000 
 
 Niagara Falls, N. Y., 
 
 4,500,000 
 
 Long Branch, N. J., 
 
 3,000,000 
 
 Terre Haute, Ind., 
 
 4,500,000 
 
 
 
 The only large city which has attempted to purify its water 
 supply by alum and mechanical filtration is New Orleans.* Louis- 
 ville, Ky., has been conducting experiments for the past two years 
 upon mechanical nitration with alum, the results of which are not 
 yet available by the public. The typhoid fever statistics of the 
 smaller cities using mechanical filters have not been compiled, but 
 the author's investigations along this line indicate a lower effici- 
 ency for rapid sand nitration with alum, than by plain slow sand 
 nitration as employed in the purification of polluted waters abroad. 
 
 " While much is claimed for the bacterial efficiency of mechan- 
 ical filters, the fact remains that the claims are not so firmly veri- 
 fied by scientific research as are those of slow sand filtration." f 
 
 REMOVAL OF IRON FROM GROUND WATERS. 
 
 From Mr. Baker's description of the mechanical filters used in 
 connection with the water-works of Asbury Park, NJ.,J the follow- 
 ing information has been drawn : 
 
 The filters were introduced to reduce the amount of iron in the 
 ground water which constitutes the supply for these works. The 
 capacity is 2,000,000 gallons per acre per day, and each pair of fil- 
 ters consists of two steel tanks, 6 feet diameter and 28 feet long, 
 
 * Nineteenth Annual Report, New Orleans Water-Works Company, 1897, p. 5. 
 f Manual of American Water-Works, New York, 1897, p. N. 
 t Proceedings New Jersey Sanitary Association, 1895, p. 92. 
 
204 THE PURIFICATION OF WATER. 
 
 of f-inch plates, with double riveted longitudinal seams. The first 
 tank of each pair of filters contains quartz or sand, and the second 
 tank contains animal charcoal. The depth of filtering material is 
 about four feet. During winter the filters are washed once every 
 24 hours, and during summer once every 12 hours. No coagulant 
 is used. Before reaching the filters the water is aerated by the 
 Pohle air lifts which are used to raise the water from the wells to 
 the receiving-tanks. The water is obtained from 7 wells, 4^-inch 
 casings, each 600 feet deep, and 2 wells, 6-inch casings, one 1,021 
 feet and the other 1,132 feet deep. The cost of the filters, includ- 
 ing foundations, is stated as 120,000. 
 
 From experiments by Dr. T. M. Drown the following net results 
 were obtained : 
 
 The iron in solution in the water of the two deeper wells was 
 1.125 and 1.1378 parts per 100,000 respectively. For the 600 
 feet deep wells, from several determinations, the iron amounted to 
 0.1791 part per 100,000. 
 
 Aeration and filtration of the water gave the following reduc- 
 tions in the amount of iron : 
 
 Filtering through sand alone, 87.9 per cent, 
 
 u H a 92.7 
 
 u u 96.9 
 
 " " 95.5 " 
 
 u a n u 97.7 " 
 
 " " " and charcoal, 95.0 " 
 
 " " " 98.0 " 
 
 n n a. u a 98.2 " 
 
 Concerning the necessity of filtration through charcoal, Dr. 
 Drown says : 
 
 " If we take the last two days only, when the pumps were working at their 
 maximum rate, the removal of iron by the sand filter was 98.3 per cent, and by 
 both filters 98.51 per cent, or only 0.21 per cent additional removal by the 
 charcoal filter." 
 
 The filters at Asbury Park have been in operation since Janu- 
 ary, 1895 ; they are of the pressure type, the water being pumped 
 under a pressure of 54.5 pounds to the consumers, with a loss of 
 3 to 5 pounds in passing the filters. 
 
MECHANICAL FILTERS. 205 
 
 The reduction of iron in ground waters by aeration, and filtra- 
 tion through sand, has been practiced for some years in Germany. 
 
 Dr. Dunbar, the director of the Hamburg Institute of Hygiene, 
 has made quite an extensive investigation of the various processes 
 employed,* from which the following facts are taken : 
 
 The earlier method, pointed out independently by Anklam and 
 by Oesten, consists of aeration of such water in long canals, or by 
 dropping the water in finely divided streams from an altitude of 
 2 meters (6.5 feet). With some waters this has given good re- 
 sults, with others it has failed almost entirely. Thus in one 
 instance the oxide of iron (Fe 2 O 3 ) was 1.10 mg. per liter in the 
 raw water, and in the treated water, showing a complete reduc- 
 tion or removal of the iron. In another instance the reduction of 
 the iron oxide was only from 24 mg. to 20 mg. per liter, showing 
 a reduction or removal of only 16 to 17 per cent of the iron in the 
 untreated water. 
 
 The Piefke method consists of the percolation of the water 
 through a chamber filled with coke. The efficiency of this method 
 is said to be increased as the coke becomes covered with the iron 
 and other substances precipitated from the water. Upon a large 
 scale this method has reduced the iron in ground water from 40 
 mg. per liter to merely a trace. 
 
 The Kronke method of iron reduction in water consists of the 
 treatment of the water first by chemicals, usually a salt of iron, 
 and then by lime. By this process all the iron can be eliminated 
 from water. The chemicals consist of 1 gram of ferric chloride 
 and 5 to 10 grams of lime to 100 liter of water. The apparatus 
 required consists of a mixing-tank, a measuring-vessel, and a filter 
 to intercept the precipitated iron. The use of this process has 
 given reductions of the iron in the untreated water as shown in 
 the first table on the following page. 
 
 The cost of the iron salts and lime required by this method, 
 according to Dr. Dunbar, varies from ^ to \ cent per cubic meter 
 (264.2 gallons) of water. 
 
 The process for reduction of iron in water by flow through a 
 
 * " On the Nature and Treatment of Ground Waters Containing Iron, etc.," Zettschrift 
 fur Hygiene, vol xxii., 1896, p. 68 et seq. 
 
206 
 
 THE PURIFICATION OF WATER. 
 
 OXIDE OF IRON IN WATER. 
 
 UNTREATED WATER. 
 MG. PER LITER. 
 
 TREATED WATER. 
 MG. PER LITER. 
 
 PERCENTAGE OF 
 REDUCTION. 
 
 8.00 
 
 0.10 
 
 98.75 
 
 9.20 
 
 0.00 
 
 100.00 
 
 10.00 
 
 0.00 
 
 100.00 
 
 1.95 
 
 0.10 
 
 94.87 
 
 24.00 
 
 0.15 
 
 99.37 
 
 bed of animal charcoal, in which the deferrization is due either to 
 oxidation in the pores of the filter, or to the influence of the calca- 
 reous constituents of the filter material, shows in some instances a 
 complete removal of the iron. Such filters diminish in efficiency 
 from day to day, but can be restored to their original capacity by 
 immersion of the animal charcoal in dilute hydrochloric acid. The 
 influence of time on. such a filter is shown in the following 
 table : * - 
 
 REDUCTION OF OXIDE OF IRON IN WATER BY FLOW THROUGH 
 A BED OF ANIMAL CHARCOAL. 
 
 DAYS OF 
 SERVICE. 
 
 ORIGINAL FE Z O 3 
 MG. PER LITER. 
 
 AFTER FLOW 
 THROUGH FILTER 
 (FE 2 3 ). 
 MG. PER LITER. 
 
 PERCENTAGE OF 
 REDUCTION. 
 
 1 
 
 . . 
 
 0.25 
 
 98.96 
 
 2 
 
 . . 
 
 0.25 
 
 98.96 
 
 3 
 
 . . 
 
 0.40 
 
 98.34 
 
 4 
 
 24.0 
 
 0.55 
 
 97.71 
 
 5 
 
 24.5 
 
 0.65 
 
 97.35 
 
 6 
 
 
 0.80 
 
 96.74 
 
 7 
 
 . . 
 
 0.75 
 
 96.94 
 
 8 
 
 . . 
 
 1.50 
 
 93.88 
 
 9 
 
 . . 
 
 2.50 
 
 89.80 
 
 10 
 
 . 
 
 1.50 
 
 93.88 
 
 11 
 
 
 1.40 
 
 94.29 
 
 The Massachusetts State Board of Health has very carefully 
 investigated the subject of iron in ground waters, and the follow- 
 ing paragraph has been taken from its Twenty-seventh Annual 
 Report : 
 
 * " On the Nature and Treatment of Ground Waters Containing Iron, etc.," Zeitschrift 
 fiir Hygiene, vol. xxii., 1896, p. 135. 
 
MECHANICAL FILTERS. 207 
 
 
 
 " Many experiments have been made at different times with a view to a 
 removal of the iron by oxidation and subsequent filtration of the water through 
 sand to remove the precipitated iron oxide. The results obtained by this 
 means have been very variable. But in general it may be said, that when the 
 iron is present in the water in small amount, say, not over 0.3 of a part per 
 100,000, the iron will separate out of the water almost completely on exposure 
 to the air for 24 to 36 hours, in the form of a rusty precipitate, which can be 
 removed entirely by filtration through sand at a rapid rate. Forced aeration 
 by filtering through sand with a current of air was found in almost all cases to 
 hasten the oxidation and separation of the iron oxide." 
 
 The performance of the iron reduction plant at Reading, Mass., 
 from a series of analyses extending from July 27, 1896, to Oct. 6, 
 1896, indicates reductions ranging from 83.16 to 100 per cent of 
 the amount originally in the water. In this instance the plant con- 
 sists of a Warren mechanical filter, used with lime, forced aeration, 
 and sulphate of alumina, for the precipitation and coagulation of 
 the iron. In one instance the iron originally in the water is stated 
 as 0.356 part per 100,000, and in the treated water as 0.020 
 part per 100,000, showing a reduction of 94 per cent.* Analyses 
 by Professor Henry Carmichael, Boston, September, 1896, gave 
 the iron oxide in the original water as 0.1070 part per 100,000, 
 and in the filtered water as 0.0107 part per 100,000, showing a 
 reduction by the Warren process of 90 per cent. 
 
 * "Removal of Iron from Ground Water," M. N. Baker, Engineering News, Nov. 26, 
 1896. 
 
208 THE PURIFICATION OF WATER. 
 
 CHAPTER XIII. 
 
 HAMBURG SETTLING-BASINS AND FILTERS. 
 
 THE following description of the water purification works at 
 Hamburg is based upon a very elaborate paper by Mr. Meyer, the 
 chief engineer.* These works were under construction at the 
 time of the cholera epidemic, September, 1892, and stimulated by 
 this awful calamity (which was largely due to the condition of the 
 water supply of the city), the work on the filters during the inter- 
 vening months was carried on day and night by the aid of electric 
 lights, and finished in May of the following year. 
 
 These works are the most modern in all proportions and ap- 
 pointments, and illustrate on a grand scale the combined effect 
 of sedimentation and filtration on the water from the River Elbe ; 
 a source of supply which will bear comparison with some of the 
 sewage-polluted rivers of this country. 
 
 The water is first pumped from the Norder Elbe into four 
 settling-basins, each 393.6 feet wide by 1,148 feet long, and 11.15 
 feet deep. The effective depth of these basins is stated as 6.56 
 feet, and the available water capacity of all as about 84,500,000 
 U. S. gallons. The usual time allowed for sedimentation is 19 to 
 30 hours, according to information furnished the author by Mr. 
 Rud Schroder, inspector of the Hamburg Water- Works. 
 
 These settling-basins are excavated in the earth, with inside 
 slopes three horizontal to one vertical. The bottom and slopes 
 are puddled, and covered with a pavement of brick or tile. Of 
 the total depth of water, 4.59 feet is below the invert of the 
 effluent pipe, and represents the space allowed for accumulation 
 of mud, silt, etc., in the bottom of the basin. The material pre- 
 cipitated by subsidence from the river water can from time to 
 
 * Das Wasserwerk der Frein und Hansestadt Hamburg, by F. Andreas Meyer, Ham- 
 burg, 1894. 
 
HAMBURG SETTLING BASINS AND FILTERS. 
 
 209 
 
 time be flushed out of the basins through 
 a 36-inch cast-iron waste-pipe, which is con- 
 trolled by a stop valve in the embankment. 
 
 " The main conduit of 
 masonry, from the settling- 
 basins on Bill warder Island 
 to the filters on the Kalte 
 Hofe, is 9,020 feet long, 8.5 
 feet diameter, until it reaches 
 the group of filters, around 
 which it is reduced to 5.25 
 feet diameter for a length of 
 885.6 feet. From this main conduit short 
 branch pipes 3.936 feet diameter lead to the 
 influent chambers of each of the filters." 
 
 " The influent chamber to each filter con- 
 tains two compartments. In the first is placed 
 a double-seat (balanced) valve connected by 
 means of a lever or walking-beam to a float in 
 the second compartment, by means of which, 
 when the water on the filter reaches the de- 
 sired elevation, the valve is closed automatic- 
 ally. The water flows upon the filter through 
 two openings in the side of the chamber, the 
 bottoms of which are at the same elevation as 
 the surface of the sand in the 
 filters (see Fig. 13). 
 
 "All the filters except 
 one are rectangular in plan, 
 this one being shaped to 
 the topography of the island. 
 All the filters are open, and 
 constructed with inside slopes, 
 two horizontal to one verti- 
 cal. Sloped embankments were used, partly 
 because the marshy soil would not support 
 walls of masonry, and partly because the ac- 
 
210 THE PURIFICATION OF WATER. 
 
 tion of the expanding ice on the paved slopes would not so readily 
 injure them." 
 
 According to the Hamburg officials, open filters have an ad- 
 vantage over closed filters in being more easily cleaned, operated, 
 and inspected. Exposure to the air can work no injury to the 
 water on the filter ; upon the contrary, it is held that the decom- 
 position of organic matter is aided by contact with the air. The 
 water in the settling-basins and filters, which becomes warmer than 
 the river water during the summer, is again cooled in the long con- 
 duits and covered reservoir.* 
 
 " The location of Hamburg, near the German Ocean, precludes 
 the probability of ice forming on the filters for long periods. It 
 should be mentioned, however, that during the winter the water in 
 the River Elbe is generally quite clear, and the intervals between 
 scrapings of the sand are therefore considerably lengthened. 
 
 " The Hamburg filters have very large dimensions compared 
 with other works, namely, 7,650 square meters (1.89 acre) each. 
 The objection to large filters heretofore has been that the clean- 
 ing of them is not so easily accomplished as are those of small 
 area. If this objection was real, there is no doubt that smaller 
 beds would be made of those now used in the Hamburg water- 
 works. The experience had with these filters seems to show that 
 the cleaning can be easily accomplished ; and there is no reason 
 why the beds should be made smaller than is necessary from an 
 economic standpoint, in order not to have an excess of filtering 
 area in reserve, or out of service during the time of scraping and 
 renewal of the sand. For this reason the size of the single filters 
 in any system of filtration should not exceed a given per cent of 
 the total filtering area. In these works, containing eighteen 
 filters, it does not appear that the unit area has been made too 
 large. By reducing the areas of filters, the walls of which are 
 sloped, a great deal of otherwise useful surface would be lost in 
 the increased length of embankments, the number of influent and 
 effluent chambers would be increased, and the land required for 
 
 * The author is informed that the water in the covered storage reservoirs at Rothenburgs- 
 ort is usually cooler by 2 degrees Fahr. in summer than the water as it is pumped from the 
 river to the settling-basins. 
 
HAMBURG SETTLING-BASINS AND FILTERS. 
 
 211 
 
 a given effective filtering area would also be increased. Finally, 
 the cost of construction and operation of the works would be 
 increased on the one hand, while, on the other, the convenience of 
 operation would be diminished. It therefore seemed to be more 
 advantageous for Hamburg to have large filter-beds ; and the expe- 
 rience of four years has demonstrated that neither in the operation 
 of the filters, nor in the bacteriological results from the filtrate, 
 can any objection be found. 
 
 "The imperviousness of the filters and embankments in the 
 more or less sandy soil of the Kalte Hofe was obtained in the fol- 
 lowing manner : 
 
 " Upon a layer of good marsh soil, 14 inches thick, was placed 
 a layer of plastic clay puddle 4 inches thick. The clay was pre- 
 pared either by ordinary clay-cutters, or by grinding in pug mills, 
 such as are used in the manufacture of tile and brick. The bot- 
 tom was paved with bricks laid flatwise in cement mortar, while 
 the inner slopes were paved with a layer of bricks on edge. The 
 upper edge of the slope paving was finished with a strip of con- 
 crete. The outer slopes of the embankments were sodded. The 
 embankments between 
 the filters were finished 
 with a gravel path, in 
 the middle of which the 
 narrow gauge railway 
 tracks for the sand cars 
 were laid. 
 
 "After the water has 
 percolated through the 
 sand and gravel to the 
 bottom of the filters, it 
 is received in collecting- 
 drains, constructed as 
 
 shown by Fig. 14. The main collecting-drain, which extends from 
 the effluent chamber of the filter, is built upon a separate founda- 
 tion in the filter-bed. It is 22 inches high, 32 inches wide in the 
 clear, with brick side walls, and cover stones of granite. Into this 
 the side drains, 7.5 inches high and 6 inches wide, are connected. 
 
 Wafer 
 
 Sand 
 
 :* 
 
 
 
 Z Meters. 
 
 1 '6' )' z' V 4-' 5' <&' V 8' &' \0' 
 
 Fig. 14. Section of Main Drain and Filtering Materials, 
 
 Hamburg, Ger. 
 
212 
 
 THE PURIFICATION OF WATER. 
 
 "The branch collecting-drains of brick, laid dry, are entirely 
 surrounded by gravel. The top of the granite slabs on the main 
 drain projects into the sand about 4 inches. The water which 
 collects in the gravel layer either flows directly into the main 
 collector, through openings in the brick side walls, or enters the 
 branch collectors through similar openings, and then flows to the 
 main collector. From the main collector the clear water enters 
 the effluent well shown in Fig. 15. 
 
 "The arrangement of the filtering materials is shown by Fig. 
 14. The water stands in the filter at about 3.6 feet above the 
 
 Chambt 
 
 ! 
 
 
 
 
 
 
 
 1 
 
 
 
 ,- H- 
 
 
 -4>~ 
 
 
 
 D 
 --- 
 
 o 
 
 -$- 
 
 .._! 
 
 i J 
 
 
 Main 
 
 
 &>//<? 
 1 
 
 f/ng 
 
 i 
 
 a 
 
 Drair 
 j 
 
 
 
 
 
 
 
 ^ 
 
 \ 
 
 
 
 
 
 
 
 V _-rt- 
 
 
 -A.. 
 
 5' 
 
 i 
 
 
 
 
 
 
 3^ J 
 
 l 
 
 Iniluent 
 Chamber 
 
 L... 
 
 ]Q 
 
 20 
 
 50 60 
 
 70 
 
 90 100 Meters. 
 
 20' 40' 60' ao' 100' 0' 
 Fig. 15. Plan of Filters, Hamburg, Ger. 
 
 sand, and must penetrate a thickness of 3.28 feet (one meter) of 
 fine sand, when the filter is first put in operation, which, upon 
 successive parings, may be diminished to a thickness of 16 inches. 
 The dirty sand is scraped off in successive thicknesses of three- 
 eighths to one-half inch, and conveyed to the sand washer to be 
 cleaned, and then takes its place again in the filter. The layer 
 of sand rests upon a layer of gravel 24 inches thick, the sizes of 
 which are so arranged that the upper finer material cannot pene- 
 trate the lower coarser material, and the latter cannot enter the 
 collecting-drains. 
 
HAMBURG SETTLING-BASINS AND FILTERS. 213 
 
 ARRANGEMENT OF FILTERING MATERIALS. 
 
 . . < Effective size, 0.30-0.34 mm. r . 
 
 Fine sand \ _ ^40 inches. 
 
 ( Uniformity coefficient, 2.0 -2.3 
 Gravel, different sizes, 24 " 
 
 Total depth of filtering materials, 64 inches. 
 
 Head of water on filter, 42 " 
 
 "To clean the filtering materials after these were brought from 
 the various natural deposits, five sand-washing machines were 
 erected at different points on the work. 
 
 " The gravel which was taken from the pits had to be freed 
 from the mud and sand attached to it, and also assorted into the 
 various sizes corresponding to the layers in the filters. This 
 double object is accomplished by means of slightly inclined cylin- 
 drical iron drums, having numerous projections on the interior. 
 The gravel being placed at the upper end, the drum is revolved, 
 while numerous jets of water are played upon the material. At 
 the lower end are placed a number of shaking sieves of various 
 degrees of fineness, one above the other, by which the gravel is 
 separated according to the commercial sizes, and then falls into 
 wagons placed under the sieves. 
 
 " The sand is taken from pits and banks. Cylindrical iron drums 
 were also used for the purpose of cleaning the sand, arranged in 
 such a way that of the raw material, when threwn upon a shaking 
 sieve, the finer particles were washed through this sieve, leaving 
 the coarser material upon it. It was then introduced into the lower 
 end of the slightly inclined, slowly revolving drum (see Fig. 29). 
 
 " By means of a number of spiral vanes upon the inside of the 
 drum, with small intervening spaces, the mixture of sand and water 
 was conveyed to the upper end of the drum, where it was dis- 
 charged into a hopper. As the sand left the drum it was again 
 played upon by a strong stream of water, which was directed 
 partly into the interior of the drum, and partly on the sand which 
 was flowing out. 
 
 " By this arrangement the sand gradually came in contact with 
 cleaner water as it rose to the upper end of the drum, while the foul 
 water found an exit from the lower end of the drum. The washed 
 sand was then shoveled into dump cars and conveyed to the filters. 
 
214 THE PURIFICATION OF WATER. 
 
 The small gravel obtained upon separation and grading of the sand 
 was washed, and used as the top layer of gravel in the filters. 
 
 During a day and night turn of the washing apparatus, the fil- 
 tering material prepared for use was about 2,614 cubic yards, for 
 which were required : 
 
 8 drums worked on stone and coarse gravel. 
 4 drums worked on smaller gravel. 
 26 drums worked on sand. 
 
 The total quantity required (for eighteen filters) amounted, in 
 round numbers, to 104,640 cubic yards of gravel and stone, and 
 248,520 cubic yards of sand. For the cleaning of the stone, 
 gravel and sand, filtered water was used exclusively, except for 
 the materials which were placed in the first filter, for which, of 
 course, filtered water was not then available. 
 
 " The regulation of the quantity of water flowing from the fil- 
 ters is accomplished in the following manner. The effluent well 
 contains several compartments, through which the water must pass 
 successively. From the main collecting-drain, the water enters 
 the first compartment, and rises to an elevation corresponding to 
 the elevation of the raw water on the filter. The water passes 
 from the first into the second compartment over an adjustable 
 weir, by means of which the quantity flowing from the filters will 
 always be the same. The dimensions of the weir are such that 
 the lower edge can be placed at a point 2.30 feet below the water 
 level of the filter. When the weir is at its highest position the 
 flow of water is checked entirely, and the filter is taken out of 
 service (see Fig. 16). 
 
 "One attendant can regulate the position of the weirs, and 
 serve ten filters if necessary. To the weir is fastened a scale, and 
 a float placed about three feet to the rear of the weir operates a 
 pointer which indicates on the scale the difference in height of 
 the water levels. The float is placed to the rear of the chamber 
 to avoid the influence of the currents of water flowing over the 
 regulating weir. 
 
 "When the difference of height is constant, i.e., when the 
 pointer covers a certain mark on the scale corresponding to the 
 
HAMBURG SETTLING-BASINS AND FILTERS. 
 
 215 
 
 quantity of water which the filter was in- 
 tended to yield, the flow over the weir will 
 be uniform for equal intervals of time. 
 Thus, if the rate of percola- 
 tion through the sand be 
 2.56 inches per hour, we can 
 deduce the height from the 
 formula, Q = a. b. h. 
 
 in which b 1.0 m. : g = 
 9.81 m., the coefficient, a, 
 being obtained from obser- 
 vation of the fall of the un- 
 filtered water when the supply valve was 
 closed. The term a was deduced as 0.503, 
 and the quantity of water per second 
 was observed to be 0.136 c. m., and (usin 
 7,650 sq. m. as the useful area of a filter), 
 we obtain h = 0.155 m. 
 
 " It was intended to measure the quan- 
 tity of filtrate by the difference in height 
 between the water levels of the two well 
 chambers separated by the weir, and to regis- 
 ter this difference in height by an automatic 
 apparatus, but attempts in this direction have 
 not given satisfactory results. 
 
 " From the second of the well chambers 
 the water passes through an 
 iron pipe into a third com- 
 partment, and from this into a 
 branch canal 2.624 feet wide, 
 and then to its respective 
 branch clear-water conduit. 
 
 " In determining the size 
 of the filter works, a velocity 
 of filtration of 2.5 inches per hour, or 5 feet 
 per day vertical, has been used as a basis 
 (this is a measure of the quantity of water 
 
216 
 
 THE PURIFICATION OF WATER. 
 
 leaving the filter when the influent valve is closed) ; and in opera- 
 tion the attempt is made to not exceed this rate of percolation, 
 and to avoid rapid changes in the rate of filtration. But variations 
 in consumption (as between holidays and week-days, or between 
 a warmer and a colder day) can Le compensated by a temporary 
 increase in the rate of filtration, as it will scarcely be possible to 
 have (at Hamburg) clear-water reservoirs of such size as will pro- 
 vide for large hourly variations in the demands for water. 
 
 It has not been proven by the experience with the Hamburg 
 filters to the present time that the filtrate is affected by such 
 variations in the rate.* In fact, the daily records of the operation 
 of the various filters indicate that, under certain circumstances, 
 there is no increase in the numbers of bacteria in the water when 
 the rate is changing. 
 
 The two following diagrams for Filters Nos. 12 and 16, which 
 have been compiled from the daily records of the Hamburg Hy- 
 gienic Institute, indicate the influence of changes in the rate of 
 filtration, on the bacterial contents of the filtered water. 
 
 k 
 
 2,000,000 
 
 1,000,000 
 
 Filter No.\2. 
 Fig. 17. Diagram Showing Operation of Filter No. 12, Hamburg, Ger. 
 
 Filter No. 12 was started in service Dec. 12, 1893, and stopped 
 for scraping of the sand Jan. 27, 1894, having run without inter- 
 ruption for 47 days. Filter No. 16 was started Dec. 6, 1893, and 
 stopped Jan. 25, 1894, having run without interruption for 51 days. 
 
 * April, 1897. 
 
HAMBURG SETTLING-BASINS AND FILTERS. 
 
 217 
 
 These two diagrams are an interesting study of the operation 
 of plain sand filters on a large scale. Referring to that for Filter 
 No. 12, it will be noted that the rate of filtration was 1,689,096 
 gallons per acre per day, for the first 24 days. On the 25th day, 
 the rate was 1,900,224 gallons per acre, for the 26th day, 1,951,972 
 gallons, and on the next day it was again 1,689,096 gallons per 
 acre ; after which, to the 47th day, the rate varied from 1,329,379 
 gallons to 1,900,224 gallons per acre per day. 
 
 During this interval of time the bacteria per cubic centimeter 
 ranged from 23 colonies on the 44th day, to 93 colonies on the 
 17th and 19th days ; never reaching the limit assigned by the 
 
 K- December 1893. ->!< January 1&94. -X 
 
 2,000,000 
 
 1,000,000 
 
 Filter No. 1 6. 
 Fig. 18. Diagram Showing Operation of Filter No. 16, Hamburg, Ger. 
 
 German Imperial Institute of Hygiene (100 colonies per cubic 
 centimeter of the filtrate), and averaging, for the whole time 
 the filter was in service, 47 colonies per cubic centimeter of the 
 filtered water. 
 
 Filter No. 16 was operated for a period of 30 days at the rate 
 of 1,689,096 gallons per acre per day, then for one day at the 
 rate of 1,900,224 gallons per acre per day, then for one day at 
 the rate of 1,951,972 gallons, and for the next two days at the 
 standard rate of 1,689,096 gallons per acre per day ; after which, 
 for the remainder of the period of 51 days, the rate varied from 
 1,329,379 gallons to 1,900,224 gallons per acre per day. 
 
 For the 51 days of service of the filter, the bacteria per cubic 
 
218 
 
 THE PURIFICATION OF WATER. 
 
 centimeter of filtrate varied from 55 colonies on the 2d day of 
 service to 6 colonies on the 25th day, the average for the whole 
 time being 21 bacteria per cubic centimeter of filtrate. 
 
 The general bacterial efficiency of the Hamburg filters is 
 shown by the following table : 
 
 SOURCE OF SAMPLE. 
 
 BACTERIA PER C. C. OF WATER. 
 
 Unfiltered water from the Elbe, 
 From the filters, 
 Average bacterial reduction, per cent, 
 
 800-3000 
 
 20- 30 
 
 98.64 
 
 Whenever a filter has reached a point in its periodical " run " 
 where it requires a head of water on the sand in excess of the 
 maximum allowed, viz., 42-44 inches, in order to obtain the 
 standard rate of delivery, it is taken out of service, the sand 
 scraped, and after observing the usual precautions in refilling, is 
 started again. 
 
 At Hamburg the drainage pipes of the filters are 20 inches 
 diameter, and controlled by an ordinary stop valve. All the drain- 
 age pipes discharge into a common masonry conduit or sewer about 
 4 feet diameter, which traverses the Kalte Hofe, and connects with 
 a pump well on the bank of the River Elbe. 
 
 (Mr. Meyer is silent upon the manner of discharging the 
 contents of this pump-well. Water drawn from the bottom of a 
 filter after it is taken out of ' service will be filtered water, and if 
 not changed in quality in passing through the drainage conduit, 
 could with safety be pumped into the clear-water conduit which 
 conveys the water from the filters to the clear-water basin at 
 Rothenburgsort. The first run of a filter after it is started in 
 operation is probably also discharged into this same drainage con- 
 duit or sewer, and collected finally in the same pump-well, from 
 which it should be pumped into the river. The double use of 
 this drainage conduit is open to criticism. In one aspect of the 
 case it indicates the waste of filtered water which might, with a 
 proper arrangement of pipes or conduits, be saved and used ; and 
 in another, the possible after pollution of water which had left 
 the filter in condition for domestic use.) 
 
HAMBURG SETTLING-BASINS AND FILTERS. 219 
 
 
 
 To refill and start a filter, water from the clear-water conduit 
 is allowed to flow backward through sluice gates in the effluent 
 well to the central collecting-drain, from this to the lateral brick 
 drains, and finally upward through the gravel and sand until it 
 stands at a depth of eight inches above the surface of the sand. 
 Further filling is then accomplished by opening the valve in the 
 influent chamber, after which the automatic float and valve in this 
 chamber, in connection with the adjustable weir in the effluent 
 chamber, regulates the head on the sand and the discharge of the 
 filter, within the limits fixed in practice. 
 
 Each filter is connected with a clear-water conduit of brick 
 masonry 8.5 feet diameter, 2,460 feet long, which lies parallel to 
 the dike of the Elbe. This conduit at one point is connected with 
 the old conduit through which the water was taken from the river 
 for the old settling-basins prior to May, 1893. The connection 
 was made by means of a side-shaft, and so arranged that, during 
 the construction of the filters, each filter could be allowed to dis- 
 charge its filtrate into the old conduit ; and after a sufficient num- 
 ber of filters were put in service to supply the whole quantity of 
 water consumed by the city, the connection between the old 
 intake from the river and the new filtered-water conduit was tem- 
 porarily closed. During September of 1893, a leak was discovered 
 in the temporary bulkhead, and the connection between the old and 
 new conduits was closed with a permanent bulkhead of concrete. 
 
 From the clear-water conduit on the Kalte Hofe, the water is 
 carried by an inverted siphon of welded steel pipe 6.56 feet dia- 
 meter, across the Billwarder Bay, to a basin or clear-water reservoir 
 on the Rothenburgsort, from which the filtered water is pumped 
 to the city. 
 
 The clear-water basin is covered with a masonry vaulting rest- 
 ing on pillars, the arches of which are coated with a layer of 
 asphalt, to exclude the water which may percolate through the 
 earth and sand which is placed above the arches, and the whole 
 covering is finally finished with a layer of sod. Drain tile is 
 placed over the arches, to carry off the seep water which may find 
 its way through the covering. 
 
 The capacity of the clear- water basin (1897) is stated to be 
 
220 THE PURIFICATION OF WATER. 
 
 6,182,280 U. S. gallons, while the average daily consumption of 
 water for 1896 is given as 31,524,080 gallons, indicating that this 
 clear-water reservoir was filled and emptied about five times each 
 day. The maximum daily consumption for 1896 was 38,407,811 
 gallons, at which time the clear-water basin contained less than 
 four hours' average consumption. The author is informed that, 
 during the summer, the temperature of the water falls about 2 
 Fahr., between the river and the clear-water reservoir, while in 
 the winter the temperature rises between these two points about 
 the same amount. 
 
 The average length of " run " of a filter between scrapings, 
 at Hamburg, is about forty days. But, as the author is informed, 
 under favorable conditions of the water from the river, during the 
 winter of 1896-1897, one filter successfully delivered, between 
 scrapings of the sand, a column of water 105 m. (344.5 feet) 
 high, which is equivalent to 112,215,438 U. S. gallons per acre, or 
 at the standard rate of percolation for these filters (1,689,096 gal- 
 lons per acre per day), represents an uninterrupted service for 66 
 days. 
 
 The average daily per capita consumption of water by Ham- 
 burg for 1896 was 50 U. S. gallons. 
 
 THE SCHRODER SAND-WASHER. 
 
 The sand-washers now in use at the Hamburg Water-Works 
 are the invention of Mr. Rud Schroder, inspector of the filters ; and 
 each set consists of seven conical boxes or hoppers of iron or steel, 
 in the lower ends of which are fitted Korting ejectors. Filtered 
 water under a head of thirty-six feet is supplied to the ejectors 
 from a manifold, while the sand is fed into the first hopper by 
 manual labor. The mixture of sand and water is carried up 
 through a vertical pipe by the action of the ejector, and dis- 
 charged into the next hopper of the series. The current of water 
 through the ejector performs two offices; one, the transportation 
 of the sand from hopper to hopper, and the other, the separation 
 and washing of the dirty sand. The dirty wash-water overflows 
 the upper edges of the hopper, and is carried off by suitable 
 
HAMBURG SETTLING-BASINS AND FILTERS. 
 
 221 
 
 troughs. Seven hoppers of the form shown in the drawing (Figs. 
 19 and 20) are found sufficient to effectually wash the sand 
 scraped from the filters, and restore it to a condition fit to go 
 into the filters again. 
 
 From the vertical pipe in each hopper a trough conveys the 
 mixture of sand and water to the next hopper of the series. By 
 Fig. 22, it will be seen that these ejector washers are arranged in 
 sets of two, each set consisting of a sand-chute, a small receiving- 
 
 Sand-Washing Plant, Hamburg, Ger. 
 
 Fig. 19 (Hopper No. 7). 
 
 Fig. 20 (Hopper No. 2). 
 
 hopper at the bottom of the sand-chute, and six larger elevated 
 hoppers. From the last hopper the washed sand is discharged 
 onto a platform, from which it is shoveled into the tram-cars. 
 The surplus water from the sand is conveyed away by troughs to 
 the rear of the platform. 
 
 These hoppers are about 2 feet 6 inches square, excepting the 
 first of the series, which is 2 feet square. The first hopper which 
 receives the sand is about 1 foot 8 inches deep, while the remain- 
 ing hoppers are about 2 feet 2 inches deep. The converging 
 chute into which the " fouled " sand is dumped is 4 feet 2 inches 
 
222 
 
 THE PURIFICATION OF WATER. 
 
 Fig. Z4. 
 Transverse Section C D 
 
 Sand-Washing Plant, Hamburg, Ger. 
 
HAMBURG SETTLING-BASINS AND FILTERS. 223 
 
 
 
 by 6 feet 6 inches at the top, and 4 feet deep. This is provided 
 with a gate and screen at the bottom, through which the flow of 
 sand to the first of the washing-hoppers is regulated. The water- 
 pipe from the manifold to the ejector is \\ inches diameter, and 
 the elevator pipe above the ejector is 3 inches diameter. A chilled 
 iron throat is screwed into the lower end of the elevator pipe to 
 resist the grinding action of the mixture of sand and water. 
 
 From Mr. Schroder's description of this apparatus,* as used 
 with the Hamburg filters, one complete set of washers and all 
 appurtenances cost about $2,400. Four sets of washers are suffi- 
 cient for the service of 18 filters of 1.89 acres each, or 34 acres of 
 filtering area. The expenditure of water per cubic yard of sand 
 washed averages 4,043 U. S. gallons, applied under a head of 36 
 feet. As stated before, filtered water only is used for sand-wash- 
 ing. This style of washer is regarded at Hamburg as being more 
 economical in labor than the drum washers employed at Berlin 
 and in the London Water-Works, although requiring about twice 
 the quantity of water for washing the sand and operating the 
 ejectors. 
 
 The capacity of one set of the Hamburg sand-washers is 
 stated by Mr. Schroder to be 4 c. m., or 5.23 cubic yards, of sand 
 cleaned per hour. The author is informed that these ejector sand- 
 washers have been in constant service since 1894, cleaning annu- 
 ally about 25,000 c. m. (32,675 cubic yards) of "fouled " sand from 
 the filters, and have during this time given entire satisfaction. 
 
 The following notes are from the operation of these Water- 
 Works for 1896 : - 
 
 Total consumption of water for all purposes, 11,506,300,000 U. S. gallons. 
 Consumed in washing sand from filters, 91,800,000 U. S. gallons. 
 Percentage of filtered water required by the Schroder sand-washers, 0.80. 
 
 (In Chap. XVII. it will be noted that the English type of sand- 
 washers, in use at Berlin, and formerly used at Hamburg, require 
 about T 4 o of one per cent of the filtered water for washing the 
 "fouled" sand scraped from the filters.) 
 
 * Zeitschrift des Vereines Deutscher Ingenieure, vol. xxxix., Hamburg, 1894. 
 
224 
 
 THE PURIFICATION OF WATER. 
 
 THE MAGER SAND-SCRAPING DEVICE. 
 
 The winters of Hamburg have been sufficiently rigorous since 
 the filters were started in service to produce some inconvenience 
 in the scraping of the sand-beds after the ice forms on the water, 
 which has occurred as early as November, and continued until late 
 in February. The following diagram, from the records of the 
 Hamburg Water- Works, illustrates the time of formation, duration 
 and thickness of the ice-cover over the niters, during the winter 
 of 1896-1897. 
 
 , 1696 ->* r 1897 
 
 Nov. ->|< December- >j<- - January >K -February > 
 
 . Zl 26 11 6 II 18 21 26 311 6 II 16 7>f 26 3l'l 6 II 16 21 26 
 
 Fig. 25. Diagram Showing Ice on Filters, Winter of 1896-1897, Hamburg, Ger. 
 
 From this diagram it will be seen that the ice began to form 
 on the filters Nov. 23, 1896, and continued until Feb. 27, 1897, 
 attaining a thickness of about 13 inches. During this time sev- 
 eral of the filters in service were cleaned by the Mager apparatus, 
 to be described. 
 
 Before the invention of the Mager sand-scraping apparatus the 
 filter-beds were cleaned by hand-dredging from the after end of a 
 scow, which was slowly moved across the water from side to side 
 of a filter, by means of a wire rope stretched from bank to bank, 
 which engaged with a whim or capstan mounted on the scow. In 
 order to conduct the work in this manner, it was necessary to first 
 cut away or break and remove the ice from that portion of the 
 water where the traverse of the scow and scrapers was to be made. 
 The scrapers were mounted on long poles, and provided with bags 
 or other receptacles for the sand scraped from the surface of the 
 
HAMBURG SETTLING-BASINS AND FILTERS. 225 
 
 bed. As rapidly as these receptacles were loaded, the scow was 
 stopped, the scrapers and accumulation of "fouled" sand lifted on 
 board and dumped. The scrapers were then put in position again, 
 and the motion of the scow resumed across the filter. 
 
 The traverse of the scow and hand-scrapers was from side to 
 side of the filter, reversing the position of the scrapers with each 
 traverse, and shifting the position of the wire rope a distance equal 
 to the width of the swath or path scraped during the preceding 
 traverse. 
 
 The hand scrapers were always worked from the after end of 
 the scow, and upon accumulating on board a load of " fouled " 
 sand, this was wheeled to the bank, and carried to the sand- 
 washers. 
 
 Fig. 26. Device for Scraping Ice-Covered Sand-Filters, Hamburg, Ger. 
 
 By successive traverses from slope to slope the whole area of 
 the sand-bed was scraped, and the filter restored to service. 
 
 The use of this apparatus required the breaking of the ice over 
 the whole water area, and involved an expense of labor and time 
 which brought about the invention of the apparatus shown in 
 Fig. 26. 
 
 The Mager device consists of a large float which impinges 
 against the under side of the ice cake, and a metal scraper hung 
 from the float by a pair of oscillating arms. Two chains, con- 
 nected to the scraper and the float, limit the oscillation of the 
 arms, and with reference to the float, regulate the depth of the 
 swath cut in the surface of the sand. To the scraper is attached 
 a bag which receives the "fouled" sand as it is cut from the sur- 
 
226 THE PURIFICATION OF WATER. 
 
 face of the bed. The float and scraper is introduced at one end 
 of the filter under the ice, and by means of two capstans, placed 
 one upon each of the two longer embankments of the filter, and 
 two wire ropes attached to the float, the float and scraper is 
 dragged from side to side of the filter without removing or break- 
 ing the ice-cover. 
 
 When the scraper has made a full traverse across the filter, by 
 pulling upon one of the two lines connected with the sand-bag, it 
 is turned inside out like a stocking, and the contents emptied on 
 the inner slope of the embankment. Upon traversing the filter in 
 the opposite direction, the scraper is reversed, and upon reaching 
 the further slope the sand is discharged by pulling upon the other 
 line and reversing the bag. In this manner the float and scraper 
 is drawn from side to side of the filter until the whole bed of sand 
 is scraped, the " fouled " sand being left upon the slopes or at the 
 edge of the sand-bed. 
 
 As each traverse is made, the capstans on the embankments 
 are moved along the filter a distance equal to the width of the 
 swath cut in the sand by the previous traverse of the scraper. 
 
 This apparatus requires only the cutting away of a narrow strip 
 of ice at each end of the filter, and at the side slopes, for the intro- 
 duction and removal of the float and scraper, for the operation of 
 the wire cables which drag the float under the ice from side to 
 side of the filter, and for handling the lines which reverse the 
 sand-bag and discharge its contents. 
 
 A comparison of the time required to scrape the sand-bed with 
 the scow and hand scrapers, with the time required by the Mager 
 float and scraper, indicates a reduced cost of labor for the latter ; and 
 a comparison of the after periods which the filter will run, and the 
 volumes of water delivered before a new scraping is required, in- 
 dicates a gain in the efficiency of the filter ; i.e., the periods of sub- 
 sequent operation are longer, and the quantities of water delivered 
 by the filter are greater, than with the apparatus previously in use. 
 
 From tables in a paper by Mr. E. Mager, descriptive of this 
 apparatus,* the following data are derived : 
 
 * Process of Cleaning the Open Filters of the Hamburg Water-Works During the Winter. 
 By Ed. Mager, Engineer, Hamburg, 1897, pp. 4, 6. 
 
HAMBURG SETTLING-BASINS AND FILTERS. 227 
 
 . BY THE OLD METHOD OF SAND-SCRAPING. 
 
 Average time required to clean one filter, 4.3 days. 
 
 Least " " " " 3 
 
 Greatest " " " " " 7 
 
 Generally 4 days were sufficient to scrape the 
 
 sand from a bed of 1.89 acres area. 
 Average length of time the filters were in ser- 
 
 vice after scraping, 15 " 
 
 Least time the filters were in service after 
 
 scraping, 4.0 " 
 
 Greatest time the filters were in service after 
 
 scraping, 29.0 " 
 
 Generally the filters were in service after scrap- 
 
 ing, 17.4 
 
 Average yield of (one) filter after scraping, 28,493,336 U. S. gallons. 
 Least yield, 7,040,402 
 
 Greatest yield, 48,725,085 
 
 Generally the average yield after scraping the 
 
 sand was, 33,588,010 
 
 Upon a second scraping of one filter by the former process, 
 6 days were required to remove the " fouled " sand, after which the 
 filter was in service for only 5 days, and the yield was 7,645,155 
 gallons. 
 
 With the Mager apparatus for scraping the sand-bed under the 
 ice, the average time required for the cleaning of three filters was 
 2.3 days each. The average period of operation of the filters was 
 17 days, and the yields of filtered water were as follows : 
 
 Average yield, 38,774,256 U. S. gallons. 
 
 Least " 24,767,165 
 
 Greatest 57,910,790 
 
 Upon a second cleaning, one of the filters required 5 days for 
 scraping the sand, with an after operation of 12 days, and a yield 
 of 19,235,873 gallons. 
 
 Generally, after cleaning the sand-bed with the Mager apparatus, 
 the yield of filtered water is from one-third to one-half as much as 
 when the same bed is laid dry and scraped with shovels in the 
 usual way. 
 
228 
 
 THE PURIFICATION OF WATER. 
 
 At present the average time required for the scraping of a sand- 
 bed of 1.89 acres with the Mager apparatus is about 40 hours. 
 
 Referring to the diagram (Fig. 25), during the interval of time 
 when the surface of the water on the niters was frozen, niters were 
 cleaned with the Mager apparatus as follows : 
 
 NUMBER 
 
 OF 
 
 FILTER. 
 
 DATE OF 
 SCRAPING. 
 
 DURATION OF 
 AFTER SERVICE. 
 DAYS. 
 
 NUMBER 
 
 OF 
 
 FILTER. 
 
 DATE OF 
 SCRAPING. 
 
 DURATION OF 
 AFTER SERVICE. 
 DAYS. 
 
 7 
 
 December 10, 
 
 
 16 
 
 January 1, 
 
 
 7 
 
 January 11, 
 
 31 
 
 16 
 
 January 22, 
 
 21 
 
 7 
 
 January 14, 
 
 
 6 
 
 January 9, 
 
 
 7 
 
 February 15, 
 
 32 
 
 6 
 
 February 21, 
 
 43 
 
 8 
 
 December 26, 
 
 
 4 
 
 January 16, 
 
 
 8 
 
 January 15, 
 
 20 
 
 4 
 
 February 7, 
 
 22 
 
 8 
 
 January 19, 
 
 
 12 
 
 January 22, 
 
 
 8 
 
 February 9, 
 
 21 
 
 12 
 
 February 24, 
 
 33 
 
 Another method which was resorted to during the past winter 
 for scraping the sand-bed is described by Mr. Schroder. The ice 
 was cut away from one-half the bed, and the water level lowered 
 until the ice-cake rested on the sand. Workmen were then put 
 upon the dry portion of the bed, and removed the "fouled" sand 
 by hand-scraping. The filter was then filled with water, the ice- 
 cake floated over to the opposite side, and the water again lowered 
 until the remaining half of the filter was laid bare. This half was 
 then scraped by hand in the same manner, after which the filter 
 was filled and put in regular service. An operation like this of 
 course requires that the temperature of the air shall be above the 
 freezing-point during the interval of time when the water is off 
 the filter. 
 
 About the first week of November, 1897, the daily newspapers 
 of the United States contained an account of an epidemic, or un- 
 warranted increase, of typhoid fever in Hamburg during the fall 
 of that year. Desirous of ascertaining if this was caused by the 
 failure of the filters to properly purify the raw Elbe water, or if 
 the increase in typhoid was traceable to other causes, the author 
 requested an explanation of the Hamburg officials ; and from the 
 letter in reply the following quotations are extracted : 
 
HAMBURG SETTLING-BASINS AND FILTERS. 229 
 
 " The investigations by the Medical Board *have shown that 
 the increase in typhoid fever during the fall of 1897 was due 
 either to the use of raw milk or unfiltered Elbe water, and there 
 was no evidence to show any connection between this rise in the 
 typhoid rates and the filtered public water supply. 
 
 "The use of the raw water for drinking-purposes is partly due 
 to the following facts : - 
 
 " For a time after our epidemic of cholera in 1892 (when natu- 
 rally the people had a dread of the unfiltered Elbe water), no cases 
 of typhoid occurred. With the lapse of time, this fear of the raw 
 river water has somewhat subsided ; and notwithstanding the warn- 
 ing signs set up at frequent intervals along the harbor against 
 using (for drinking or dietetic purposes) the unfiltered river water, 
 some people associated with the river interests are reckless enough 
 to use this water. 
 
 " Excepting such cases as were due to raw (unsterilized) milk, 
 it may therefore be of interest to you to know that all other cases 
 of typhoid fever have been stated by our Medical Board as being 
 derived from the use of raw river water in the harbor, and not 
 from the use of our filtered water, which remains up to date of 
 excellent quality." 
 
230 THE PURIFICATION OF WATER. 
 
 CHAPTER XIV. 
 
 THE FILTERS OF THE BERLIN WATER-WORKS. 
 
 THE original filters of the Berlin Water- Works at the Stralau 
 station, built in 1855-1856, were uncovered, and as described by 
 Mr. Kirkwood,* consisted of six beds, with an area of about 4.86 
 acres, or 0.81 acre to each filter. The filtering materials consisted 
 of small boulders, gravel, and sand. No lateral drains were used to 
 convey the filtered water to the central drain, the boulders at the 
 bottom of the bed performing this office. 
 
 The arrangement of the filtering materials was as follows : 
 
 Fine sand at top of bed (effective size, 0.35 mm.), 18 inches. 
 
 Coarse sand, 12 " 
 
 Coarser sand, 6 " 
 
 Gravel under the sand, and boulders at bottom of filter, 22 " 
 Total depth of filtering materials, 58 inches. 
 
 Depth of water on filters, 54-60 " 
 
 According to Mr. Kirkwood, ice from 15 to 20 inches in thick- 
 ness had formed over the filters during " long and severe winters ; " 
 and as a protection to the filter walls, the ice-cake was broken 
 around the edges " by workmen appointed to that duty." This 
 difficulty with the ice, and the impossibility at that time of properly 
 cleaning the sand-beds in winter, led to the adoption of covered 
 filters in the Lake Miiggel Works, to be described. As originally 
 constructed, the filters at the Stralau station were open ; but since 
 1893 these filters have been covered. 
 
 The new water-works of Berlin are located on the north shore 
 of Lake Miiggel,f a branch and enlargement of the River Spree, 
 
 * Filtration of River Waters, by J. P. Kirkwood, 1869, p. 112. 
 
 f The Filtration of the Miiggel Lake Water Supply, Berlin, by Henry Gill, Institution of 
 Civil Engineers, London, 1895, p. 14. 
 
THE FILTERS OF THE BERLIN WATER-WORKS. 
 
 231 
 
 about 12 miles from the center of the city. *Lake Miiggel, so- 
 called, is 2.90 miles long and 1.43 miles wide, with a depth over 
 the greater part of the area of 26 feet. 
 
 Unaltered Water 
 Inlet Pipe 
 
 Pilfered Water 
 Out-let Pipe 
 
 Washout Pipe 
 
 Fig. 27. Plan of Filters at Lake Miiggel, Berlin, Ger. 
 
 The works are designed to supply 47,280,000 gallons of water 
 per day, and contain 44 filters, each of an area of 0.576 acres,. or a 
 total sand surface of 25.344 acres, divided into four groups of 11 
 
232 THE PURIFICATION OF WATER. 
 
 filters each. It is assumed in these works that 3 filters of each 
 group will at all times be out of service for cleaning and renewal of 
 the sand, or be held in reserve ; hence T 8 T of the total filtering 
 capacity only will be available. The estimated rate of filtration 
 is 2.448 gallons per square foot of sand surface per hour, or 
 2,559,237 gallons per acre per day. The available filter area is 
 assumed to be 18.432 acres, and the daily capacity as 47,171,856 
 gallons per day. 
 
 The filtering materials are arranged as follows : 
 
 Fine sand at top of bed (effective size, 0.35 mm.), 24 inches. 
 
 Gravel, 12 
 
 Boulders, 12 " 
 
 Total depth of filtering materials, 48 inches. 
 
 Head of water on filters, maximum, 31.5 
 
 Head of water on filters, minimum, 3.6 " 
 
 (Mr. Gill estimates the voids or water space in the compacted sand as 
 J of the whole mass.) 
 
 The rate of discharge from a filter is a nearly constant quan- 
 tity ; and to effect this with an increasing head on the sand-bed, the 
 water flows from the effluent chamber to the clear-well through a 
 submerged orifice, the head over which is maintained at a uniform 
 height by the adjustment of a sluice gate, placed in the division 
 wall of the effluent chamber, between the sluice chamber and the 
 weir chamber. The adjustment of the gate from day to day serves 
 to maintain the difference of level between the water on the filter 
 and the water in the sluice chamber, to obtain the desired rate of 
 percolation through the sand, and the proper head on the sub- 
 merged orifice in the weir chamber. (See Fig. 28.) 
 
 By this device (the invention of Mr. Gill) any variation in the 
 demand for water can neither increase nor diminish the rate of 
 flow through the filters, and will only lower or raise the level of 
 water in the clear-well. 
 
 The Muggel filters are covered with groined arches, the sup- 
 porting piers of which are placed 14.37 feet center to center in 
 each direction. At the center of each arch over four piers, an 
 opening is placed, which admits of a thorough lighting up of the 
 bed of sand for the purpose of cleaning. 
 
THE FILTERS OF THE BERLIN WATER-WORKS. 
 
 233 
 
 According to the experience at Zurich, the cost of operating is 
 lower for the closed filters than for the open filters,* while the re- 
 verse seems to be true at Berlin ; for the closed filters are said to 
 entail a cost of $10.00 f per million U. S. gallons, the highest cost 
 for any European filtration works from which reports have been 
 obtained by the author. Aside from the special difficulties due to 
 the formation of ice on the open filters in climates like that of Ber- 
 lin, the cost of scraping, removing, and renewing the sand should 
 
 Filter 
 
 Showing Level of Water 
 in Sluice Chamber. 
 
 Fig. 28. Plan of Regulating Chamber. 
 (Gill on the Filtration of the Miiggel Lake Water-Supply.} 
 
 be the least with open filter-beds, and why the closed filters at Zu- 
 rich (after omitting the charge for breaking the ice) should cost 
 less for operating than the open filters, requires some explanation. 
 At Berlin the clear-water reservoir is also covered with a 
 masonry vaulting; and the whole work, as described by Mr. Gill, 
 is of the most substantial kind. 
 
 * Water Supply of Ziirich, Preller, p. 37. 
 
 f Said to include interest and sinking-fund charges. 
 
234 THE PURIFICATION OF WATER. 
 
 The new water-works at Lake Miiggel were planned to supply 
 a population of 1,700,000 with an allowance of 27.5 U. S. gallons 
 per capita, which would indicate an approximate present pumpage 
 and filtration of 46,750,000 U. S. gallons of water per diem. 
 
 From an elaborate description by the late Mr. Henry Gill of the 
 method for operation of the filters at Lake Miiggel, the following 
 resume is taken. 
 
 The filter is started in service by filling from below with filtered 
 water. The water is allowed to percolate slowly upward through 
 the bed of sand in order to displace the air and fill all the voids 
 between the sand-grains. In filling a filter the influent and efflu- 
 ent gates are closed, and the water drawn back through an inde- 
 pendent valve and pipe from the clear-well. As soon as the water 
 has risen 4 inches above the bed of sand, the influent gate is 
 opened, and further filling is cautiously conducted with unfiltered 
 water. After a head of 1.6 feet above the sand has been attained, 
 the unfiltered water is quickly run on the filter until the full ope- 
 rating head is reached. Mr. Gill recommends that a filter be filled 
 at a rate of not more than 4.7 inches per hour, to avoid disturbance 
 of the sand. 
 
 After a filter has been filled, it should be rested with the water 
 over the sand-bed for 24 hours, in order that the pores of the sand- 
 bed may be partially closed by sedimentation ; and in cases when 
 this length of rest is inadmissible, and the filter must be put in 
 service earlier to maintain the supply of filtered water, it should 
 be brought very gradually up to its normal work.* 
 
 Sudden variations in head or pressure on the sand-bed should 
 be avoided, to prevent injury to the film of silt and the products 
 of bacterial action at the surface of the sand. After the normal 
 rate of filtration has been attained, the deposit on the surface of 
 the sand increases from day to day, and the effective head neces- 
 sary to obtain the normal discharge of water from the filter will 
 also have to be increased by adjustment of the sluice gate in the 
 
 * In the Journal of the Sanitary Institute, October, 1895, p. 387, Professor Percy Frank- 
 land says : 
 
 " It is of importance to hasten the formation of the surface slime ; and to this end the water 
 should be run onto the filter, and left there undisturbed for twelve hours before filtration is 
 actually commenced." 
 
UNIVERSITY 
 
 THE FILTERS OF THE BERLIN WATER-WORKS. 235 
 
 
 
 effluent chamber. With these filters, according to Mr. Gill's rules, 
 when the difference of water level on the sand-bed and in the 
 sluice chamber reaches 1.64 feet, the filter must be taken out of 
 service. 
 
 Upon taking a filter out of service the influent and effluent 
 valves are closed, and the water level lowered to the layer of gravel, 
 or to the floor of the filter, no water being left in the bed of sand. 
 It is desirable, each time a filter is taken out of service, to thor- 
 oughly aerate the sand-bed. The upper surface of the sand is 
 pared with shovels, the cut in the surface not exceeding 0.4 inch. 
 Care should be observed to avoid taking off a thicker layer of the 
 "fouled" sand. The scraped sand is gathered in heaps in the 
 center of each vault, and carried in barrows to the sand-house for 
 washing and storage. 
 
 Mr. Gill thinks that a closed filter, after cleaning, should be 
 exposed to the atmosphere for several days (excepting in winter, 
 when the temperatures are below the freezing-point), and thor- 
 oughly ventilated before it is started again. 
 
 " Fouled" sand from a filter is washed and stored until required 
 to renew the thickness of bed in the filter. In all cases the origi- 
 nal thickness is reduced to 16 inches. After the last paring has 
 been taken from the surface, the bed is filled with washed sand 
 until the original thickness is obtained. 
 
 After sand has been scraped from a filter and taken to the 
 washing-machine, it should be so thoroughly washed that a sample 
 stirred in a beaker of distilled water will produce no turbidity. 
 
 The sand- washers used at Berlin (Fig. 29) are of the revolv- 
 ing-drum type,, the kind originally in use at Hamburg, and which 
 were discarded there for the ejector washers described in Chap- 
 ter XIII. 
 
 In the operation of these washers, the rate at which the mate- 
 rial is worked through the drum will depend upon the speed of 
 rotation, whil'e the quantity of water supplied to the drum is regu- 
 lated by a tap or valve. By varying the speed of rotation and the 
 flow of water, a thorough washing of the material, no matter how 
 foul it may be, can be accomplished by the time the sand reaches 
 the discharge end of the drum. 
 
236 
 
 THE PURIFICATION OF WATER. 
 
 A circular weir at the inlet end can be raised or lowered, and 
 thus increase or diminish the volume of water retained at all times 
 in the drum. It is advisable (especially in summer) to wash the 
 
 scraped sand as soon as it comes from the filters, and store it ready 
 for future use. 
 
 The earliest recorded comparison (see Chapter XI.) of steril- 
 
THE FILTERS OF THE BERLIN WATER-WORKS. 237 
 
 ized sand, and sand washed but not sterilized, was made at the 
 Stralau station of the Berlin Water-Works, with the result that 
 the best nitrate invariably was obtained from a bed of washed, 
 unsterilized sand.* 
 
 The water from Lake Miiggel is very variable in bacterial 
 contents, having sometimes as many as 6,000 colonies per cubic 
 centimeter, and at other times so few as 200 colonies per cubic 
 centimeter. In the operation of the niters it is the aim to keep 
 the bacterial contents of the filtrate within the German standard, 
 i.e., 100 colonies per cubic centimeter of water; and seldom do the 
 numbers of bacteria exceed 90 per cubic centimeter in the filtered 
 water, while counts as low as 40 per cubic centimeter are often 
 made. Not considering the time when the bacterial contents of 
 the lake water is very low, the general reduction of bacteria by 
 the niters is nearly 99 per cent. 
 
 * Lake Miiggel Water Supply, Gill, p. 9, 10. 
 
238 THE PURIFICATION OF WATER. 
 
 CHAPTER XV. 
 
 THE FISCHER FILTER AND ANDERSON PURIFIER. 
 
 THE FISCHER PLAQUE FILTER. 
 
 THIS is an invention of Mr. Fischer, Director of the water- 
 works of Worms, Germany, where it has been in operation for 
 four years past, and consists of hollow plates or bricks about one 
 meter (40 inches) square and 20 cm. (8 inches) thick, with 5 cm. 
 (2 inches) of space in the middle of the plate, which gives an 
 effective thickness of filtering-plate of 3 inches. These plates, 
 or plaques, are made of a mixture of clean sharp sand and finely 
 pulverized glass, obtained from the waste of glass-works, broken 
 bottles, etc. This mixture, when fused, may be given any form 
 desired, and upon cooling forms a porous mass through which 
 water may be filtered under pressures depending primarily upon 
 the density of the material. The head under which this form of 
 filter works at Worms is given in the Consular Report* as 3 to 4 
 feet. 
 
 From a drawing which accompanies the Report, it appears that 
 the hollow brick, or plaque as it is called, is made up of two solid 
 plates, 40 inches square and 3 inches thick, bolted together on a 
 frame of metal (?), with which the plates make water-tight joints, 
 and leaving a water space or cell between the plates 2 inches in 
 width. These hollow plates or bricks are set on edge in two tiers, 
 as shown by the drawing, in a suitable water-tight tank or res- 
 ervoir, with a water space of 3 or 4 inches between adjacent pairs 
 of plates. The reservoir is then filled with water until a head is 
 obtained sufficient to secure the desired rate of filtration through 
 the plates. 
 
 The water passes through the 3-inch thickness of plate from 
 the tank to the cell inside, from which, by suitable pipes, it is 
 
 * Advance Sheets of U. S. Consular Reports, February, 1897. 
 
THE FISCHER FILTER AND ANDERSON PURIFIER. 239 
 
 drawn off to the 
 clear-water reser- 
 voir. The suspend- 
 ed matter in the 
 water is intercept- 
 ed at or near the 
 outer surface of the 
 plates ; and when 
 the pores become 
 so plugged as to re- 
 duce the capacity of 
 the filter to a rate 
 of delivery at which 
 it becomes unprof- 
 itable to operate it, 
 the water is drawn 
 from the tank, and 
 by reversing the 
 current the filtered 
 water is caused to 
 pass from the cen- 
 tral cells outward 
 through the pores of 
 the plates, and the 
 accumulated s u s - 
 pended matter in- 
 tercepted at the 
 surface is washed 
 away, and flushed 
 from the plates and 
 the tank by a hand 
 hose. 
 
 The principle of 
 filtration is the same 
 as that employed in 
 the Berkefeld and 
 Pasteur type of fil- 
 
240 THE PURIFICATION OF WATER. 
 
 ters, and the method of reversing the current to wash the filter 
 the same as is employed in the mechanical filter. 
 
 The original sand filter at Worms contained 13,000 square feet 
 of filtering surface, and filtered at the rate of 792,510 U. S. gal- 
 lons per day, equivalent to 2,655,700 gallons per acre per day ; 
 while a battery of 500 of these Fischer plates or hollow bricks 
 is said to have yielded the same amount of filtered water as the 
 sand filter. 
 
 Estimating the effective area of one face of a plaque at one 
 square meter, and of both faces at two square meters, or 21.528 
 square feet, then 500 such plaques (of which 30 are shown in Fig. 
 30, each pair of plates being bolted together making a hollow 
 brick) would contain 10,764 square feet, and the rate of percolation 
 through the 3 inches of porous material would be 73.6 gallons 
 per square foot per day, equivalent to a vertical rate of 9.8 feet per 
 day. 
 
 (It is stated in the Report that the estimated cost of a sand 
 filter of 13,000 square feet of area was 130,000, which would make 
 the cost per acre more than 1100,000. This figure seems to be in 
 error ; for nothing approaching it in cost has heretofore, within the 
 author's knowledge, been reported. Open filters in series, includ- 
 ing clear-well and accessories upon an elaborate plan, can be con- 
 structed in this country within a cost of $40,000 per acre, and 
 estimating concrete coverings at $11,000 per acre, the cost of 
 covered filters may be as low as $51,000 per acre, which is about 
 one-half the cost assigned in the Report for the sand filters in use 
 at Worms prior to the introduction of the Fischer filter.) 
 
 The Fischer filter cost $9,600, or about $12,000 for 1,000,000 
 gallons of daily capacity. 
 
 The Report states, " From a long series of analyses and careful 
 observations made by the sanitary authorities at Worms, it appears 
 that the efficiency of the two systems of filtration, which are there 
 worked side by side, are practically identical, so far as regards 
 their effect upon the chemical purity of the water ; but the percent- 
 age of bacteria left by the Fischer process is somewhat greater 
 than is left by the sand filter, when clean and in good working 
 condition." 
 
THE FISCHER FILTER AND ANDERSON PURIFIER. 241 
 
 The porosity of these Fischer plaques is doubtless greater 
 than the porcelain tubes of the Pasteur-Chamberland filter, through 
 which bacteria are known to grow within a few days after sterili- 
 zation ; and since sterilization of these sand and glass plaques is 
 not practicable, only washing with a reversed current of filtered 
 water, there is danger of the same deterioration of quality of 
 filtrate which has often been observed by the continuous use 
 (unsterilized) of the Pasteur tubes. 
 
 THE ANDERSON REVOLVING IRON PURIFIER. 
 
 The following description of this device and its mode of opera- 
 tion is taken from a recent publication by the Anderson Purifier 
 Company entitled Water Purification : 
 
 " This process consists in passing the water while on its way to the settling- 
 beds through a wrought-iron cylinder (Fig. 30 ), supported horizontally on 
 hollow trunnions forming the inlet and outlet to the cylinder. This is kept 
 in continual slow rotation, and contains a charge of metallic iron in small 
 pieces. The iron is continually lifted and showered down through the water 
 by means of scoops fixed within the cylinder. The speed of rotation of the 
 machine is about 6 feet per minute at the periphery. The water is passed 
 through at the rate of from a third to a fifth of the capacity of the cylinders 
 per minute, thus keeping the water in contact with the iron for from three to 
 five minutes, according to the quality of the water. The cylinders are made in 
 various sizes; for example, a machine 18 feet long and 5 feet in diameter is 
 capable of treating nearly a million gallons per day, and is charged with about 
 2 tons of any sort of scrap iron, one of the most convenient forms being punch- 
 ings from boiler plates. 
 
 " This churning with scrap iron causes the water to take up a small quantity 
 of iron, from a tenth to a fifth of a grain per gallon, which, in precipitation, 
 effects the purification of the water. 
 
 " On leaving the cylinder these particles of iron are in the form of ferrous 
 hydrate ; but as the water is immediately exposed to the influence of the air, this 
 becomes quickly changed to ferric hydrate, which is precipitated in particles 
 more or less coarse according to the nature of the water under treatment. On 
 leaving the cylinder the water is passed into a settling-bed, or simple troughs, 
 in which the iron is completely oxidized by exposure to the air, and in which 
 the precipitate immediately settles. 
 
 " The action of the ferric hydrate on all impurities in the water is one of 
 coagulation, the formation of a precipitate in the water tending to throw out 
 of solution the dissolved organic substances. This explanation of the action 
 of the iron process upon the organic impurities of a water applies equally well 
 
242 
 
 THE PURIFICATION OF WATER. 
 
 to its action upon microbes. Experience shows that the microbes become 
 entangled in the precipitate, and either subside with it to the bottom of the 
 settling-tank, or remain upon the surface of the filter/' 
 
 After extensive experiments had been made with this process 
 upon the sewage-polluted water of the River Seine at Boulogne- 
 
THE FISCHER FILTER AND ANDERSON PURIFIER. 243 
 
 sur-Seine, with very gratifying results according to the report of 
 Dr. Miquel, it was adopted by the Compagnie Generale des Eaux, 
 for the supply of the suburbs of Paris. The process has been 
 made a part of the works at Choisy-le-Roi, Nogent-sur-Marne, and 
 Neuilly-sur-Marne, and is proportioned for the treatment (at all 
 the stations) of 18,500,000 U. S. gallons per diem. 
 
 From Dr. Miquel's bacterial tests of the performance of this 
 process at Boulogne-sur-Seine, the following averages for a period 
 of six months, February to July inclusive, 1893, are taken : 
 
 COLONIES PER C. C. OF WATER. 
 
 Unfiltered water from River Seine, 396,000 
 
 Filtered water, 1,702 
 
 Percentage of reduction, 99.57 
 
 The water of the River Vanne, at -the same time, contained 
 1,110 colonies per cubic centimeter. This is a very pure water 
 from protected mountain sources, 107 miles distant, and in Paris 
 is regarded as the standard for dietetic water. 
 
 The method pursued by Dr. Miquel and other French workers 
 in bacteriology is calculated to show the bacteria per cubic centi- 
 meter of a water sample about ten times as high as the method 
 employed in Germany, England, and America ; and for comparison 
 with our statements of the bacterial counts from various waters 
 his figures should be divided by this number, which will give about 
 the following results : 
 
 BACTERIA PER C. C. 
 
 Seine water (unfiltered), 39,600 
 
 Seine water (filtered), 170 
 
 Vanne water, 111 
 
 The special merit of the Anderson process is found in its 
 ability to increase the rate of precipitation of the suspended mat- 
 ter, including bacteria, without the use of chemicals as a coagulant, 
 the same result being accomplished by the "ferrous hydrate," 
 formed by the contact of the iron particles with the water, which 
 upon aeration is precipitated as a " ferric hydrate." 
 
 According to the circular from which the above information was 
 drawn, the expense of operating the small plant, including the 
 
244 THE PURIFICATION OF WAITER. 
 
 filters, at Boulogne-sur-Seine, was about $1.50 per million U. S. 
 gallons. The process is not in use in any English water-works, 
 although tried at one time under unfavorable conditions at the 
 water-works of Worcester.* Of this test Dr. Dupre says: 
 
 "1. The revolving purifier process, judged merely from a chemical point 
 of view, has been a considerable success as regards at least 7 out of the 11 
 fortnightly samples examined ; and if the process could be conducted in such 
 a manner that all the filtered water equaled these, there would be nothing left 
 to desire ; while from a bacteriological point of view it has been eminently suc- 
 cessful in practically every case. 
 
 " 2. From a sanitary point of view most of the samples of filtered water 
 are open to no objection. 
 
 "3. The process, as hitherto worked at Worcester, does not effect any 
 very noticeable reduction in the color of the water whenever there is much 
 peaty matter present. 
 
 " 4. From a sanitary point of view the presence of peat is not, however, a 
 serious evil. 
 
 " 5. Similar results might no doubt have been obtained by means of sand 
 filtration only. To obtain them in this way it would, however, be necessary to 
 increase the present filtering area by at least 50 per cent, since the rate of fil- 
 tration should then not exceed 4 inches per hour; whereas the present rate of 
 supply cannot be kept up under a rate of at least 6 inches per hour, and even 
 then no provision would be made to supply the place of filters thrown out of 
 work for cleaning." 
 
 The author's experiments with a small Anderson purifier (on 
 the Ohio River water) have given as averages of several tests 
 from 86.13 to 97.28 per cent reductions of the bacteria in the raw 
 water ; but, as stated in Chapter III., this purifier, and especially 
 the filter used in connection with it, were not calculated to favor 
 the process, and should not be weighed against the more elabo- 
 rate experiments of Dr. Miquel, 1893, on the process as used at 
 Boulogne-sur-Seine. 
 
 The claims by the manufacturers for the Anderson rjrocess of 
 treatment before the water enters the subsiding reservoirs are : 
 
 " 1. Filtration, after the water has been purified by means of the revolving 
 purifier process, is carried on at about twice the customary speed, thus effect- 
 ing a saving of about half the area of filter surface required. 
 
 * Report of Dr. A. Dupre, London, November, 1892. 
 
FILTERS PROPOSED FOR CINCINNATI. 245 
 
 " 2. The saving thus effected much more than counterbalances the cost of 
 the revolving cylinders. 
 
 " 3. The purification is much more thorough and much less liable to ac- 
 cidental disturbance, and removes a greater percentage of microbes. 
 
 "4. The working cost is low, as the iron employed is very cheap, and 
 with efficient settlement the cost of filter cleaning is very small." 
 
 The cost of a revolving purifier plant, including all usual con- 
 nections, is stated at $5,000 per million gallons of daily capacity. 
 A series of cylinders, and the usual connections and appurtenances 
 (not including the filters), required to treat 20,000,000 gallons of 
 water daily, would thus cost $100,000. 
 
 According to Mr. E. Devonshire of the Anderson Purifier 
 Company, the cost of plain sand filters abroad is $15,000 to 
 $20,000 per million gallons of daily capacity, while with the re- 
 duced filter area required by the Anderson process, these figures 
 are reduced to $9,000 and $12,000 per million gallons of daily 
 capacity ; and he estimates the average cost of a combined purify- 
 ing and filtering plant at $15,500 per million gallons of capacity 
 per day. It is stated by the company that the cost of treatment 
 by this process, including the care of the sand filters, "will not 
 exceed $2.00 per million gallons." 
 
246 THE PURIFICATION OF WATER. 
 
 CHAPTER XVI. 
 
 FILTERS PROPOSED FOR CINCINNATI. 
 
 THE water supply of Cincinnati has for years been in a deplo- 
 rable condition, and different measures for relief have been proposed 
 at various times during the past thirty years. As early as 1865 
 Mr. James P. Kirkwood proposed settling reservoirs and plain sand 
 filters for the treatment of the Ohio River water before it was 
 supplied to the consumers. This method of water purification, 
 with such modifications as the intervening time has suggested, 
 was recommended by the Commission of Engineers appointed to 
 report plans and estimate of cost for extension and betterment of 
 the city water-works.* The plans embrace subsidence in large 
 reservoirs for four days previous to the delivery of the water to 
 the filters. 
 
 The premises and conclusions upon the matter of water puri- 
 fication, as set forth in the Report of the Engineer Commission, 
 abridged and corrected for the present purpose, were as follows : 
 
 " Experiments indicate that subsidence for four days will remove 
 from the Ohio River water a very large percentage of the suspended 
 matter, and relieve the filters of that part of the work which is 
 chiefly concerned in the clarification of the water. The effect of 
 this will be to cause the filters to pass a larger quantity of water 
 per unit of area between successive parings or cleanings of the 
 sand." 
 
 " Much of the work now required of the filters abroad will be 
 accomplished in the subsiding reservoirs ; and by a fair division of 
 the work between the subsiding reservoirs and the filters, relying 
 upon the former largely for clarification and improvement of the 
 color, and upon the latter wholly for the reduction of the bacteria 
 
 * This commission reported March 20, 1896. 
 
FILTERS PROPOSED FOR CINCINNATI. 247 
 
 (and finer suspended matter), better results can be had in the 
 quality of effluent and economy of operation than by filtration 
 alone." 
 
 " The subsiding reservoirs have been designed for ready cleans-- 
 ing from the silt and other suspended matter in the water which 
 will be deposited upon the bottom and slopes, and are so arranged 
 in unit capacity that at all times at least 250,000,000 gallons, and 
 usually 300,000,000 gallons, cf sedimentation capacity will be in 
 service." 
 
 " The filters were designed for a total capacity of 66,000,000 
 gallons per day, and a least effective capacity of 60,000,000 gallons 
 per day. The net aggregate area of water and sand surface is 
 22 acres, allotting two acres to each of the eleven filters. The 
 estimated rate of delivery is 3,000,000 gallons per acre per day." 
 
 "To obtain the highest quality of effluent with the maximum 
 allowable rate of filtration, regulators will be used on both the 
 inflow and outflow pipes, limiting the head on the sand-bed and 
 the loss of head between the water on the filter and the level of 
 water in the effluent chambers to such measures as may be found 
 to give the most satisfactory results in practice." 
 
 The subsiding reservoirs are six in number, and each has a 
 capacity of 50,000,000 gallons when filled to a depth of 30 feet. 
 The bottom dimensions are 705 feet by 210 feet, with dimensions 
 at the full water-line of 855 feet by 360 feet. The top width of 
 embankment has been fixed at 20 feet, with inside slopes 2J hori- 
 zontal to one vertical, and outside slopes 2 horizontal to one verti- 
 cal. The bottom and inside slopes are to be covered with 2 feet 
 of puddle, over which will be a pavement of concrete 6 inches 
 thick. The top of the embankment will be paved with concrete 
 and small broken stone rolled in place, to form a foot-walk and 
 driveway around and between the reservoirs. 
 
 The dimensions of sand-bed and water surface of each filter 
 are 220 feet wide by 400 feet long. The depth of the filter, from 
 the top of coping to the concrete floor, is 11 feet. The filters 
 have been planned with masonry walls, vertical on the inside and 
 battered by offsets on the outside. Under the bottom of the filter 
 a layer of puddle 12 inches thick has been shown, and over this 
 
248 
 
 THE PURIFICATION OF WATER. 
 
 puddle is placed a concrete floor 6 inches thick. The walls are 
 started on a course of puddle 12 inches thick, with a broad footing, 
 
 Fig.32b. Transverse Section. 
 F/grs. 37 anrf 32. Proposed Filter Bed for the City of Cincinnati, 0., 1896. 
 
 and around the walls puddle of varying widths will be packed up 
 to the level of the ground. 
 
FILTERS PROPOSED FOR CINCINNATI. 249 
 
 Each filter has two acres of sand and water surface,* and is 
 provided with two main drains laid to a grade of 6 inches in 200 
 feet, each main drain being graded from the center of the length 
 of the filter chamber to the effluent chambers at the ends of the 
 filter, to collect the water from one-fourth the area of the filter, 
 and discharge this right and left to the effluent chambers." The 
 dimensions of each section are, therefore, 110 feet by 200 feet. 
 
 One filter of 2 acres sand area is divided into four parts or 
 sections of \ acre each, from which the filtered water is collected, 
 and delivered to the four effluent chambers ; and any variation in 
 the quality of filtrate supplied by each \ acre of the filter can be 
 detected by proper tests of the water at the effluent chambers. 
 The sections, Figs. 32 a and 32 b, show the method of construction 
 proposed. The excavation will be carried to such bottom eleva- 
 tion as will provide for the filling between and for the embank- 
 ments around the filters, with due allowance for shrinkage in 
 volume of material by rolling and the action of the elements. 
 
 The concrete floor and masonry side and end walls constitute 
 the basin or reservoir for the reception of the main and lateral 
 drains and the filtering materials. Each section of the floor is a 
 shallow trough, with a general grade toward its respective effluent 
 chamber. The regulator on the influent side of the filter consists 
 of a 30-inch balanced valve, and a metal float which closes the 
 influent valve when a head of 4 feet over the sand layer is reached. 
 Upon the effluent side the yield of the filter is conducted to the 
 clear-well through a 24-inch pipe, controlled by a balanced valve 
 and float, which limits the delivery of each \ acre of the filter to 
 1,500,000 gallons per day. 
 
 " The main drains are built of brick, with portholes in the 
 three upper courses to receive the water from the small lateral 
 drains, and are covered with close jointed stone slabs 3 inches 
 thick. The walls of the main drains are 12 inches thick, and 
 rest on a concrete foundation 6 inches thick." 
 
 The main drains are 2 feet wide and 2 feet high in the clear, 
 two of which are provided for each acre of filtering surface. 
 
 " The lateral drains are of vitrified salt-glazed tile, with butt 
 
 * Compare Hamburg Filters, Chap. XIII. 
 
250 THE PURIFICATION OF WATER. 
 
 joint of arched section with flat bottoms, and perforated on the 
 top and sides. The inside dimensions are 6 inches wide and 
 8 inches high. These are laid on the concrete floor to a grade 
 of 3 inches in 52.5 feet. The lateral drains are spaced 11.8 feet 
 center to center. 
 
 " The filtering materials are arranged as follows : 
 
 Fine graded sand (at top of filter), 30 inches. 
 
 Coarse sand, 15 " 
 Small gravel, 6 " 
 
 Coarse " L5 " 
 
 Total depth of filtering materials, 66 inches. 
 
 Depth of water over sand, 48 " 
 
 " Each filter is provided with one influent and four effluent 
 chambers ; and each chamber is provided with an automatic regu- 
 lating-valve to control the depth of water over the sand-bed, and 
 regulate the rate of flow from the filters to the clear-well. Each 
 filter is supplied through a 30-inch branch pipe, connected with a 
 48-inch supply main. Each branch pipe is provided with a stop- 
 valve to shut off the flow to the filter when it is out of service 
 and being cleaned. Provision also is made for the draining of the 
 water to such level below the surface of -the sand-bed as may be 
 desired, or to empty the filter of water altogether. 
 
 From the influent chamber two lines of 20-inch cast-iron pipe 
 pass right and left across the ends and down the longer sides of 
 the filter, from which short pieces of 8-inch cast-iron pipe de- 
 liver the unfiltered water on the filter-bed. These branch pipes 
 are placed 40 feet center to center. 
 
 The author believes that the influent pipes should enter the 
 filter a few inches (not more than one foot) above the sand, to 
 avoid the necessity of a complete refilling of a filter with filtered 
 water, and the disturbance of the sand surface by the fall of water 
 from the influent pipes while filling a filter to the standard level. 
 
 " The clear-well is planned as a masonry structure, with walls 
 vertical on the inside and battered by offsets on the outside, and 
 is started on a layer of puddle 18 inches thick, over which is 
 placed a layer of concrete 6 inches thick. Outside the walls pud- 
 
FILTERS PROPOSED FOR CINCINNATI. 251 
 
 die of varying widths will be rammed up to 'a level with the 
 ground. The clear-well inside has a length of 1,180 feet and a 
 width of 148 feet, giving a net area of 4 acres, which, with a 
 depth of 15 feet, contains 20,000,000 gallons, or one-fourth of the 
 daily capacity of the high-service pumping-engines. 
 
 " Much thought has been bestowed upon the problem of open 
 and closed filters for Cincinnati, and due consideration has been 
 given to the practice of filter construction abroad. In latitudes 
 where the winters are rigorous it is essential that the filters be 
 covered to secure good results. 
 
 " In temperate climates, like that of London, all the filters are 
 open. In the extreme climates of St. Petersburg, Warsaw, and 
 Koenigsberg the filters are covered to avoid the danger due to a 
 complete freezing over of the water on the sand-bed, and more 
 especially to prevent freezing of the sand when the filter is taken 
 out of service. 
 
 "The filters of Berlin (a city in a climate nearly like that of 
 Cincinnati), are covered, while the latest filter works of Germany, 
 those of Hamburg, are of the open type. 
 
 " The normal temperature of the winter months should govern 
 in this matter ; and a comparison of the temperatures of the three 
 winter months for the past eleven years for Cincinnati, with the 
 mean January temperatures of Berlin and Hamburg, are given in 
 the following table : 
 
 MEAN NORMAL WINTER TEMPERATURES. 
 CITY. DECEMBER. JANUARY. FEBRUARY. 
 
 Cincinnati, 36.75 30.66 34.27 
 
 Berlin, . . 31 
 
 Hamburg, . . 31 
 
 " From this it appears that the mean January temperature of 
 Cincinnati is about the same as that of the German cities noted ; 
 but of the eleven years embraced in the average for Cincinnati, 
 seven had mean January temperatures below the freezing-point. 
 
 " In the light of th.e long and valuable experience of the other 
 German cities in the matter of filter construction and operation, it 
 is difficult to conceive how Hamburg could have made a mistake 
 
252 THE PURIFICATION OF WATER. 
 
 in a matter apparently so easy of solution as the covering or non- 
 covering of its niters. Altona, adjoining Hamburg, and subject 
 to the same winter climate, had used open filters for thirty-two 
 years before Hamburg built its filters ; and although some com- 
 plaint had arisen in Altona against open filters, it does not seem 
 that this was strong enough to cause the use of covered filters in 
 Hamburg. 
 
 The commission decided to recommend filters without cover- 
 ings, with a provision in the report for the vaulting of the filters, 
 should experience demonstrate the necessity of these. Informa- 
 tion from Hamburg, received since the Report was submitted, 
 indicates that the covering of filters in climates similar to Cincin- 
 nati is not essential to satisfactory results, either in the quality of 
 filtrate or in the management of the filters during the winter. 
 
 The project of water purification for Cincinnati contemplates 
 eleven open filters, each of two acres of sand and water surface ; 
 and estimating in the usual manner for engineering structures, 
 the cost of the filters, including all necessary pipes, valves, regu- 
 lators, etc., is $65,146.50 for one filter. The cost of clear-well of 
 masonry construction was $162,696.90, and the total cost for the 
 filtering-works was estimated as follows : 
 
 ORIGINAL PLAN FOR CINCINNATI FILTERS. 
 
 11 filters, 2 acres of filtering area each, at $65,146.50, $716,611.50 
 Clear-well, 20,000,000 gallons capacity, 162,696.90 
 
 $879,308.40 
 
 Add 10 per cent for sand-washing and conveying ma- 
 chinery, contingencies, etc., 87,930.84 
 45.06 acres of land, at $150.00, 6,759.00 
 Total for 11 filters, clear-well, and all appurtenances, $973,998.24 
 Cost per acre of filtering area, 44,272.65 
 Cost per 1,000,000 gallons of estimated capacity, 14,757.55 
 
 Subsequent to the report of the commission the author investi- 
 gated the cost of these filters, if constructed after the Hamburg 
 plan, with sloped walls of earth instead of masonry, the sand sur- 
 face to remain the same as before ; viz., two acres, influent and 
 effluent chambers of masonry, and all distributing and collecting 
 
FILTERS PROPOSED FOR CINCINNATI. 253 
 
 9 * 
 
 pipes and channels of the same construction as before, from which 
 
 the following resume is drawn : 
 
 AMENDED PLAN FOR CINCINNATI FILTERS 
 
 11 filters, 2 acres of filtering area each, at $53,960, $593,560.00 
 
 Clear-water reservoir, 20,000,000 gallons capacity, at 
 
 $3,000 per million, 60,000.00 
 
 $653,560.00 
 
 Add 10 per cent for sand-washing and conveying ma- 
 chinery, contingencies, etc., 65,356.00 
 52.5 acres of land, at $150.00, 7,875.00 
 Total for 11 filters, clear-water reservoir, and all appur- 
 tenances, $726,791.00 
 Cost per acre of filtering area, 33,036.00 
 Cost per 1,000,000 gallons of estimated capacity, 11,012.00 
 
 In this estimate the clear-water reservoir is also considered as 
 a plain earthen reservoir with sloped walls, paved with concrete 
 six inches thick, same as filter basins. 
 
 The estimates of cost for a system of plain sand filtration for 
 Cincinnati were made from plans prepared with unusual care in 
 view of the novelty of the proposition to filter 60,000,000 gallons 
 of river water per diem, and it is not known that any feature of 
 successful filtration was omitted in the plans or overlooked in the 
 estimate. The prices for materials and construction adopted in 
 the detailed estimates are really higher than the prices prevailing 
 at this time (1897), and the author is confident that entirely 
 satisfactory and durable works can be constructed within the 
 estimates given. 
 
 With reference to the great cost of the filter works at Berlin, 
 Mr. Gill's plans have been studied very carefully ; and aside from 
 the fact that he has included in the cost of the purification works 
 the cost of certain features of the Lake Miiggel works, which in the 
 author's opinion are not strictly chargeable to the filters, but should 
 be charged to the pumping-works, the whole work was conducted 
 upon a very costly scale, scarcely justified even in permanent works 
 of public water supply. 
 
 (During a long experience with public works the author has 
 seen much money wasted in certain details of construction which 
 
254 THE PURIFICATION OF WATER. 
 
 have been due to inexperience or perverted judgment upon the 
 part of the constructors. Thus reservoirs, thoroughly constructed, 
 complete in every essential, and as durable as such structures may 
 be, have been completed by some engineers at a cost of $2,500 per 
 million gallons of available capacity, while in other situations equally 
 as favorable for this class of works, the engineer has succeeded in 
 using up over $4,000 per million gallons of reservoir capacity, with 
 no material gain in the quality of the finished structure. Point 
 lace on the legs of a pair of overalls, or the sleeves of a machinist's 
 jacket, would add nothing to the utility of these garments, while 
 costing many times as much as the garments themselves. In like 
 manner the author has seen works overloaded with trimmings 
 which attract the eye, while obviously lacking in some of the 
 essentials for convenient service and durability. In the construc- 
 tion of filters and appurtenances the same extravagance which is 
 often displayed in other engineering structures may enter into 
 these with no benefit whatever to the works themselves.) 
 
 The manner in which the filters and appurtenances for the 
 Hamburg Water- Works were forced to an early completion by 
 working night and day, in order to avoid the possible return of 
 cholera in 1893, manifestly increased the cost of labor, and possibly 
 of materials, over the cost of what might have been obtained with 
 slower and more deliberate construction ; and yet these works are 
 very complete in every essential of modern filter construction, and 
 cost a trifle over one-half, per acre of effective area, than that of 
 the filters in the Berlin Water-Works. The Berlin filters are cov- 
 ered ; but allowing $13,000 per acre for concrete vaulted covers, 
 then the Hamburg filters would have cost but two-thirds as much 
 as the Berlin filters. 
 
 (The people are always ready to condemn the prodigality of a 
 private spendthrift, and how much more justly may we condemn 
 the public spendthrifts who, through ignorance and arrogance, 
 presume to squander the contents of the public purse.) 
 
COST OF FILTERS AND FILTRATION. 255 
 
 CHAPTER XVII. 
 
 COST OF FILTERS AND FILTRATION 
 
 So many variable conditions enter into the cost of construct- 
 ing a system of water filters, that any figures given in a work of 
 this kind must be accepted rather as suggestions than estimates 
 which can be used with safety in any locality. The topography 
 and nature of the ground upon which filters are to be constructed, 
 the local prices of labor and materials, the nearness and quality 
 of available filtering materials, the character throughout the year 
 of the water to be filtered, the necessity of previous sedimenta- 
 tion with certain waters, the quality of filtrate to be obtained, and 
 many other obvious conditions, must be taken into consideration 
 in determining and designing a system of filters on a large scale 
 for public water supply. The same ingenuity and judgment in 
 the use of available locations and materials of construction are 
 to be taken advantage of in the building of works of filtration, as 
 in other engineering structures. Successful filtration seems to be 
 more dependent upon management than on the construction of the 
 works, and plain construction, such as we find in the purification 
 works of Hamburg, with the skill and vigilance there displayed in 
 the management of the filters, will meet all practical requirements. 
 
 PHILADELPHIA, PA. 
 
 Mr. Hazen, in a Report to the Woman's Health Protective 
 Association,* estimates the cost of filters there (omitting value 
 of land) as given in the first table on page 256. 
 
 These prices include settling-basins, low-lift pumps, filters and 
 clear-wells, and the pipes, valves, and regulators required to co-n- 
 nect the filters in service. In all instances noted above, the maxi- 
 
 * A Practical Plan for Sand Filtration in Philadelphia, 1896. 
 
256 
 
 THE PURIFICATION OF WATER. 
 
 STATION. 
 
 DAILY AVERAGE 
 CAPACITY, 
 GALLONS. 
 
 GROSS COST. 
 
 COST PER AVERAGE 
 MILLION GALLONS. 
 
 Belmont, 
 Queen Lane, 
 Cambria, 
 Frankford, 
 
 14,000,000 
 26,000,000 
 60,000,000 
 20,000,000 
 
 $ 317,000 
 587,000 
 1,578,000 
 389,000 
 
 $22,643.00 
 22.577.00 
 26,300.00 
 19,450.00 
 
 mum capacity of the filters is 50 per cent above the average, and 
 a statement of cost upon the total or maximum daily capacity of 
 filters will be as follows : 
 
 STATION. ' 
 
 COST PER MAXIMUM 
 MILLION GALLONS. 
 
 STATION. 
 
 COST PER MAXIMUM 
 MILLION GALLONS. 
 
 Belmont, 
 Queen Lane, 
 
 $15,095.30 
 15,051.30 
 
 Cambria, 
 Frankford, 
 
 $17,533.30 
 12,966.60 
 
 From Mr. Shedd's estimates on filters for Providence, R.I., the 
 following costs are obtained : 
 
 TYPE OF FILTER. 
 
 DAILY CAPACITY, 
 GALLONS. 
 
 GROSS COST. 
 
 COST PER MILLION GAL- 
 LONS OF CAPACITY. 
 
 Mechanical, 
 Plain sand, 
 Proposal on plain sand, 
 
 15,000,000 
 
 
 
 $281,000 
 208,000 
 200,000 
 
 $18,733.30 
 13,866.60 
 13,333.30 
 
 According to Mr. Hazen,* covered filters on the European 
 model will cost $70,000 per acre, or allowing for a rate of filtra- 
 tion of 2,500,000 gallons per acre per day, the cost per million 
 gallons of daily capacity will be 828,000 ; but upon comparison 
 with careful estimates by the author, on open and covered filters, 
 for the city of Cincinnati, this price is excessive. 
 
 Open filters, not including clear-well, sand-washing and con- 
 veying machinery, and land, if built on favorable ground, should 
 cost, in the vicinity of Cincinnati, as follows: 
 
 With vertical masonry walls, per acre of filtering area, $32,573.30 
 With earthen embankments, and slopes paved with 
 
 concrete, per acre of filtering area, $26,980.00 
 
 * Filtration of Public Water Supplies, p. 120. 
 
COST OF FILTERS AND FILTRATION. 257 
 
 Estimates furnished the author by an expe*rt in concrete con- 
 struction indicate that concrete coverings for above filters will 
 cost about 111,000 and 113,000 respectively per' acre of effective 
 sand surface ; making the cost of covered filters as follows : 
 
 COVERED FILTERS. 
 
 With vertical masonry walls and concrete coverings, 
 
 per acre of filtering area, $43,573.00 
 
 With earthen embankments, slopes paved with con- 
 crete and coverings of concrete, per acre of filter- 
 ing area, $39,980.00 
 
 In his Report to the city of Albany, Mr. Hazen modifies his 
 estimate of cost somewhat.* Here he gives the cost of eight fil- 
 ters (0.70 acre each) at $251,000, to which may be added $10,000 
 for piping about filters, etc., making the cost of covered filters 
 $46,607 per acre, a price which agrees more nearly with the re- 
 sults of the author's estimates of cost of construction. 
 
 The clear-well, or reservoir, to equalize the delivery from a 
 system of filters, should have a capacity equal to \ or \ the maxi- 
 mum daily yield of filters ; and this, if constructed as a plain reser- 
 voir with paved inner slopes, will cost from $2,500 to $3,500 per 
 million gallons of capacity. 
 
 Assuming 10 filters, open pattern, with earthen walls and 
 paved inner slopes, of the dimensions given in the description 
 of the Cincinnati filters (Chapter XVI.), the total cost of filters 
 and clear-well, exclusive of land, should not exceed : 
 
 10 filters, 2 acres each, $539,600.00 
 
 Add 10 per cent for sand washing and conveying 
 
 machinery, and contingencies, 53,960.00 
 
 Clear-water basin, of a capacity of 20,000,000 gal- 
 lons, at $3,000 per million, 60,000.00 
 
 Total, exclusive of land, $653,560.00 
 
 Cost per million gallons of daily capacity (allowing 
 one filter to be always out of service, and 2,500,000 
 gallons average daily rate of filtration per acre), $14,523.55 
 
 * Report on Filtration of Water Supply, Albany, 1897, p. 27. 
 
258 THE PURIFICATION. OF WATER. 
 
 In the estimates for plain sand filters for the proposed exten- 
 sion and betterment of the Cincinnati Water-Works, the clear- 
 well was designed as a masonry structure of 4 acres area and 
 17 feet deep, and estimated to cost $163,000, or at the rate of 
 $7,409 per acre of filtering area. 
 
 Estimating on open filters and clear-well of masonry construc- 
 tion, which may be desirable or necessary in some locations : 
 
 10 filters, 2 acres each, $651,466.00 
 
 Add 10 per cent for sand washing and conveying 
 
 machinery, and contingencies, 65,146.60 
 
 Clear-well, 148,180.00 
 
 Total, exclusive of land, $864,792.60 
 
 Cost per million gallons of daily capacity (allowing 
 one filter always to be out of service, and 2,500,000 
 gallons average daily rate of filtration per acre), $19,217.60 
 
 COST OF FILTERS, INCLUDING CLEAR-WELLS AND ALL APPURTENANCES. 
 
 Berlin, covered (Lake Miiggel),* $68,000 per acre. 
 
 Berlin, uncovered (Stralau), 48,570 " " 
 
 Hamburg, uncovered, 38,857 " " 
 
 The Berlin covered filters, as will be observed upon reference 
 to the description of these in Chapter XIV., are of a very costly 
 construction ; and some of the appurtenances included in the cost 
 are usually found in existing water-works to which filtration may 
 in the future be applied. The Hamburg filters are of the most 
 recent and modern construction, and approach more nearly the 
 estimated cost for open filters in series in this country. 
 
 From Mr. Preller's paper on the Zurich Water-Works,! the 
 cost of covered filters was $70,857 per acre, and for the open fil- 
 ters $46,464 per acre, prices which are higher than those of other 
 works in Germany. These prices are stated to include only the 
 filter basins and filtering materials, and possibly the pipes, valves, 
 and conduits necessarily included in the construction of the filter 
 basins and influent and effluent chambers. 
 
 * Mr. Gill states the cost of the covered filters at Lake Miiggel as 15,000 per acre, equal to 
 $72,750.00. Mr. Anklam, superintendent of these works, furnishes the prices given above, 
 f Proceedings Institution of Civil Engineers, London, 1892. 
 
COST OF FILTERS AND FILTRATION. 259 
 
 The cost of the Lawrence, Mass., filter was*$26,000 per acre ; 
 and this price it is understood covers all the work in construction 
 of the filter and its connection with the previously existing filter 
 gallery, which then became the clear-well of the filter. 
 
 The city of Ashland, Wis.,* recently constructed three small 
 plain sand filters, each of \ acre area, at a gross cost of $40,178.00. 
 It was estimated that the extra cost of these filters due to local 
 difficulties was $5,367.00, and the net cost under ordinary condi- 
 tions was assumed at $34,811.00 for one-half acre. These filters 
 are covered with masonry instead of concrete vaulting, which also 
 materially increased their cost. Two of these filters have sand of 
 an "effective size" 0.27 mm. and "uniformity coefficient " 1.9 ; 
 and one has sand of an "effective size" 0.40 mm. and a "uni- 
 formity coefficient " 1.6. The sand in all beds is 4 feet thick. 
 
 RATES OF FILTRATION FOR PLAIN SAND FILTERS. 
 
 The rates of filtration per acre per day, as practiced in different 
 cities, are given as follows : 
 
 At the time of Mr. Kirkwood's visit to Europe the daily average 
 rate of percolation was 3,920,400 gallons per acre, equal to 12 feet 
 vertical per day. Mr. James Simpson, engineer of the Chelsea 
 and Lambeth Water- Works (the pioneer in sand filtration), adopted 
 a standard rate of 86.4 gallons per square foot (3,763,584 gallons 
 per acre per day), corresponding to a vertical rate of percolation of 
 11.55 feet per day. The present average rate for the London fil- 
 ters is about 1,800,000 gallons per acre per day, corresponding to 
 a vertical rate of percolation of 5.52 feet. 
 
 The filters of the New River Works sometimes reach a rate 
 as high as 3,136,320 gallons per acre per day,f while the rate at 
 which it becomes no longer profitable to operate a filter is placed 
 by Mr. Hervey of the West Middlesex Works at 1,303,000 gallons 
 per acre per day ; and when the rate of percolation reaches 2 inches 
 per hour the filter is taken out of service, the sand-bed scraped, 
 and the filter started for another period of useful work. 
 
 * Engineering News, Nov. 25, 1897, p. 338. 
 
 t According to Mr. E. L. Morris, engineer of these works. 
 
260 THE PURIFICATION OF WATER. 
 
 The Hamburg rate is 1,700,000 gallons per acre per day, cor- 
 responding to a vertical rate of 5.22 feet. Mr. Gill proposed for 
 the Miiggel Lake filters for Berlin a vertical rate of 8 feet per 
 day, equal to 2,606,630 gallons per acre per day. The rate at 
 Zurich is 5,850,000 gallons per acre, equal to a vertical rate of 
 17.95 feet per day. 
 
 The rate proposed for a sand filter to be used in connection 
 with the Marston Lake Water Supply for the city of Denver, was 
 195,500,000 gallons per acre per day, corresponding to a vertical 
 rate of percolation of 600 feet.* (There is a suggestion in the 
 paper which describes these filters that they were to be operated 
 with a coagulant, but a late report on them indicates that this was 
 abandoned, if ever used.) A filter devised for Tacoma, Washing- 
 ton, is said to work at rates of 22,000,000 to 44,000,000 gallons 
 per acre per day, corresponding to vertical rates of 67.5 and 135.0 
 feet per day.f Plain sand filtration cannot be continuously con- 
 ducted at such rates as these with any improvement in the quality 
 of the water ; and they are here mentioned in order that the con- 
 trast between these rates and the rates which long experience 
 abroad has sanctioned may be impressed upon water-works offi- 
 cials, with the hope that such works of water purification as may 
 be attempted in this country, instead of showing an utter disre- 
 gard of fundamental principles, will, if it is possible, be constructed 
 and operated upon plans which will yield even better results than 
 the works found in the cities of Europe. 
 
 Mr. Hazen, in estimating upon the cost of filters and filtration, 
 employs a rate of 2,500,000 gallons per acre per day, correspond- 
 ing to a vertical rate of percolation of 7.68 feet. 
 
 From the reports of the Massachusetts State Board of Health, 
 satisfactory results, both chemically and bacterially, in the filtrate, 
 were had with rates of filtration as high as 7,500,000 gallons per 
 acre, corresponding to a vertical rate of 23 feet per day. 
 
 The Ashland (Wis.) plain sand filters, for the year ending 
 February 28, 1897, were worked at an average rate of 2,180,064 
 gallons per acre per day, equivalent to a vertical rate of 6.69 feet. 
 
 * Transactions American Society of Civil Engineers, vol. xxxi., pp. 158-60. 
 ! Transactions American Society of Civil Engineers, vol. xxxv., p. 44 et seq. 
 
COST OF FILTERS AND FILTRATION. 261 
 
 
 DURATION OF SERVICE OF FILTERS. 
 
 The period or interval of time between cleanings or renewal 
 of the sand surface of a filter has a direct bearing on the cost of 
 filtration. Obviously the clearer the water and the lower the rate 
 of percolation the longer will be the interval of service. With a 
 given condition of the water as it comes to the filter the capacity 
 can be stated in millions of gallons filtered between cleanings, and 
 the capacity divided by the average rate of percolation per day will 
 give the number of days of filter service. 
 
 Thus, a filter which, between parings of the sand-bed, will 
 deliver 60,000,000 gallons per acre, at an average rate of perco- 
 lation of 2,000,000 gallons per acre per day, will have a period 
 of operation of 30 days. A filter which will deliver 100,000,000 
 gallons per acre, between parings of the sand-bed, at an average 
 rate of percolation through the sand of 2,500,000 gallons per acre 
 per day, will have a period of operation of 40 days. 
 
 The period of operation for the London filters ranges from 
 30 to 40 days, depending upon the condition of the water as it is 
 drawn from the River Thames or River Lea, and the time allowed 
 for subsidence in the storage reservoirs before the water is put on 
 the filters. From the evidence taken by the Royal Commission 
 on Metropolitan Water Supply the period of operation was given. as 
 short as 21 days in one instance, and as long as 70 days in another. 
 One witness stated that some of the filters of his works had been 
 in service over forty years, without any attention being given to 
 them other than the scraping, washing, and replacing of the cleaned 
 sand in the beds from time to time. 
 
 The Hamburg filters, omitting the short periods which have 
 been mentioned as occurring during the winter, have worked for 
 periods of 47, 51, and 66 days. However, the usual period of ser- 
 vice at Hamburg is about 40 days. At Zurich the covered filters 
 are reported to have an average period of service of 50 days, while 
 the open filters require cleaning every 40 days of use. In these 
 works the sand-bed is scraped successively until the remaining 
 thickness is reduced to 12 inches. 
 
 At Berlin the filters are scraped after about every 40 days of 
 
262 THE PURIFICATION OF WATER. 
 
 service ; and once in four years the whole bed of sand is taken out, 
 washed, and replaced in the filter. 
 
 During the experiments conducted by Mr. Weston, with small 
 plain sand filters at Providence, at rates of percolation less than 
 30,000,000 gallons per acre per day, the periods of operation 
 ranged from 30 to 50 days, while at the higher rates of percolation 
 the period of operation was about 20 days. 
 
 The general practice by the London water companies, in 
 restoring a filter to service from time to time, is to scrape off 
 about | inch of the clogged sand until a minimum thickness 
 (varying with the different companies) of the sand-bed is reached, 
 whereupon the whole bed of sand is readjusted in position. The 
 sand remaining after the last scraping is then taken out of the 
 filter, and replaced by the sand previously scraped from the bed 
 and washed, above which the other sand in the filter is spread, 
 scraped off, and washed in due time. By this method the whole 
 bed of sand at long intervals is scraped off, taken to the washer, 
 and returned to the filter. Thus during one complete cycle of 
 "filling" and "scraping" of the sand-bed, the whole body of sand 
 will be rotated, the lower sand coming to the top of the filter, and 
 the previous top sand going to the bottom, thus avoiding the 
 probability of converting the lower portion of the sand-bed into 
 a favorable soil for the cultivation of bacteria. 
 
 As stated in a previous chapter, the filters of some, if not all, 
 the London water-works are generally scraped and the sand de- 
 livered on the banks of the filter by contract, the price paid being, 
 as stated by Mr. W. B. Bryan, engineer of the East London Water 
 Company, X5, or about $25.00 per acre. 
 
 The period of operation of the sand filter at Lawrence, Mass. 
 (1895), was about 27 days, and at each scraping of the sand-bed 
 | inch was taken off and washed. The sand is washed by ma- 
 chinery. During 1895, 1,500 cubic yards of "fouled" sand were 
 scraped from the filter and washed, at a cost of 68 cents per 
 cubic yard, or 81.02 per million gallons of water filtered. About 
 \\ cubic yards of sand were scraped per million gallons of water 
 drawn from the filter. At Hamburg about two cubic yards of 
 sand are scraped off the filters and washed per million gallons 
 
COST OF FILTERS AND FILTRATION. 263 
 
 of water filtered. There the cost per cubic yard of scraping and 
 washing sand is considerably less than the cost of washing alone 
 at Lawrence. This is partly to be accounted for by the much 
 larger quantity of sand scraped and washed at Hamburg, and by 
 the lower cost of common labor. 
 
 In the report of the Ashland (Wis.) filters mentioned on 
 page 259, it is stated that the cleaning of the sand of one bed 
 ( acre) for the year ending February 28, 1897, consumed 
 clay, and cost 18.50, making the cost of removing and cleaning 
 the sand per acre $51.00. The total cost for cleaning and renew- 
 ing the sand for one year was $899.37, during which time the 
 filters delivered 397,860,000 gallons of water, with a cost of 
 $2.26 per million gallons filtered. While the cost per acre for 
 cleaning and renewing the "fouled" sand is very high even for 
 small filters like these, the cost per million gallons of water filtered 
 is correspondingly low, and suggests the probability of a poor 
 quality of filtrate. 
 
 LOSS OF WATER IN CLEANING FILTERS. 
 
 The cost of washing sand at Berlin is stated at 2,020 gallons 
 of water per cubic yard ; while at Hamburg, with the ejector 
 washers, the consumption of water is said to be 4,040 gallons per 
 cubic yard of sand. At Zurich the cost of washing sand by 
 machinery is given at 17 \ cents per cubic yard, but no mention is 
 made of the quantity of water required, while the cost of washing 
 and placing the sand in the filters is reported as 46 cents per 
 cubic yard. This is for the. new sand, and the price doubtless is 
 larger than for washing and replacing the " fouled " sand scraped 
 from the filters. 
 
 In Chapter XII., the statement is made upon the authority of 
 Mr. E. B. Weston that the filtered water required to wash the 
 sand-bed of the Morison mechanical filter during the Providence 
 tests, and the water run to waste after the filter was started, 
 represented about eight per cent of the water filtered, leaving 
 thus 92 per cent available for consumption. 
 
 The data upon the proportion of filtered water from mechani- 
 
264 THE PURIFICATION OF WATER. 
 
 cal filters actually available for consumption is rather meager, and 
 some of that which we have not very exact. Quoting from Mr. 
 Baker's paper on the use of mechanical filters by certain cities of 
 New Jersey,* the mechanical filters at Long Branch require about 
 5 per cent, while the filters at Asbury Park require about 10 per 
 cent, of the total pumpage for washing the sand-beds. At Key- 
 port, " Filtered water is used in washing. When the filters had 
 been in operation only some two months, it was stated that 15 per 
 cent of the water pumped was required for washing, but that the 
 contractor had promised to reduce this." 
 
 In a circular published by one of the manufacturers of me- 
 chanical filters, it is stated in one instance that 3.97 per cent of 
 the total pumpage was used for washing the sand-bed. 
 
 With the mechanical filters at Lorain, Ohio, it is stated that 
 5.22 per cent of the filtered water is used in washing the sand.f 
 
 At Hamburg the water for sand-washing (1896) represented 
 less than one per cent of the total delivery by the filters, while at 
 Berlin less than one-half per cent of the filtered water is lost in 
 washing the sand scraped from the sand-beds. 
 
 COST OF FILTRATION. 
 
 The cost of operating sand filters abroad, according to the 
 statement of the officials, are so various and widely different as to 
 suggest that some of these include items of expense not necessa- 
 rily connected with filtration per se, but which are embraced in the 
 ordinary expenses of water-works operation, while some are given 
 so ridiculously low as to raise a suspicion of error in the opposite 
 direction. 
 
 The cost per million gallons of water treated (not including 
 interest and sinking-fund charges, and omitting the charge for re- 
 moval of ice) for the Lawrence, Mass., filters for 1895 was 84.10 ; 
 and estimating interest charges on $65,000 at 5 per cent, and sink- 
 ing-fund payments invested at 4 per cent for 40 years, the total 
 cost per million gallons for that year was 7.69. Taking account 
 
 * M. N. Baker, Proceedings New Jersey Sanitary Association, 1895, p. 84 et seq. 
 t Ohio Sanitary Bulletin, October, 1897, p. 115. 
 
COST OF FILTERS AND FILTRATION. 265 
 
 of the cost of clearing the filter of ice, and including interest and 
 sinking-fund charges, the total cost per million gallons of water 
 filtered was $10.34. 
 
 On page 121 of Mr. Hazen's book on The Filtration of Pub- 
 lic Water Supplies, the cost per million gallons for treatment of 
 8,000,000 gallons per day, including interest and sinking-fund 
 charges at six per cent (no time of redemption given), is esti- 
 mated at $12.50. Omitting interest and sinking-fund charges, 
 the cost is figured at 85.30. In his report to the Woman's 
 Health Protective Association of Philadelphia, he puts the cost, 
 with previous sedimentation, at $3.50 per million gallons ; while at 
 Albany, with preliminary sedimentation, he estimates the cost at 
 $2.50 per million gallons, and without preliminary sedimentation, 
 at $3.50 per million gallons. (These prices do not include inte- 
 rest and sinking-fund charges.) 
 
 The cost of filtration, not including interest and sinking-fund 
 charges, at Zurich, is stated as 61 cents for the covered filters, 
 and 94 cents for the open filters, per 1,000,000 U. S. gallons. 
 The average cost for both open and closed filters, including inte- 
 rest and sinking-fund charges, is deduced from Mr. Preller's notes 
 as $6.71 per million U. S. gallons filtered. 
 
 A very careful estimate of all items of expense entering into 
 the operation of the filters proposed for Cincinnati, with due 
 allowance for loss of sand in handling and washing, superin- 
 tendence, daily laboratory work, and depreciation of such portions 
 of the apparatus as is subject to wear, based upon 60,000,000 
 gallons of water treated daily, gave $3.50 per million gallons, 
 exclusive of interest and sinking-fund charges ; although $4.00 per 
 million gallons was used in estimating the probable cost of filtra- 
 tion in the report on these works. 
 
 The cost of filtration in any instance (omitting interest and 
 sinking-fund charges) will depend very largely upon the manage- 
 ment of the filters. In London the cost per million U. S. gallons 
 ranges from $1.15 to $2.00, and probably averages less than $1.50 
 per million U. S. gallons. 
 
 In the review of the sand filters at Poughkeepsie, N.Y.,* the 
 
 * Manual of American Water Works, 1889-90, p. 175. 
 
266 THE PURIFICATION OF WATER. 
 
 cost per million gallons of water filtered is given as $1.32, a price 
 which indicates that these filters were not then worked with a 
 view to the high quality of filtrate obtained in works abroad. 
 
 Investigations during the present year (1897) of filter practice 
 in one of the larger cities of Germany, indicates a cost there of 
 less than $1.20 per million U. S. gallons of water filtered, for the 
 scraping of the sand-bed, transport of the sand to the washers, 
 washing the "fouled" sand, and finally returning the sand to the 
 filter bed. This price would apply only to large works, in which 
 the construction of filters and all appurtenances were modern, and 
 when the management was the best. Allowing 100 per cent addi- 
 tional for other labor about the filters ; renewal of the sand lost 
 in handling and washing ; deterioration of barrows, trucks, sand- 
 washers, etc., and for supervision ; and increasing this cost by 50 
 per cent for similar works in this country, the cost, not including 
 interest and sinking-fund charges on filters and apparatus, nor re- 
 pairs of the filters proper, should not exceed $3.00 per 1,000,000 
 gallons of water passed through the filters. 
 
 Assuming a cost of $4.00 per million gallons of water filtered, 
 interest charges on the cost of constructing open filters at 4 per 
 cent, and payments to sinking-fund for 40 years invested at 3<V 
 per cent, the total cost per million gallons of water filtered should 
 not exceed $6.37. Allowing for a consumption of 100 gallons 
 per capita per diem, the annual cost per capita will be 23 cents. 
 
APPENDICES. 
 
268 
 
 APPENDIX A, 
 
 APPENDIX A. 
 
 TYPHOID FEVER STATISTICS 
 
 OF THE PRINCIPAL CITIES OF THE UNITED STATES AND EUROPE. 
 
 Compiled from the Official Reports of Health Departments, January, 1897. 
 DEATH RATE PER 1OO.OOO OF POPULATION LIVING. 
 
 City. 
 
 Source of Supply. 
 
 1890. 
 
 POPULA- 
 TION. 
 
 DEATH 
 BATE. 
 
 New York, N.Y., 
 
 Imp'd water from Croton and Bronx Rivers, 
 
 1,705,980 
 
 21 
 
 Chicago, 111., 
 
 Lake Michigan, 
 
 1,208,664 
 
 83 
 
 Philadelphia, Pa., 
 
 Schuylkill and Delaware Rivers, 
 
 1,046,964 
 
 64 
 
 Brooklyn, N. Y., 
 
 Impounded water from driven and open wells, 
 
 853,945 
 
 26 
 
 St. Louis, Mo., 
 
 Mississippi River, 
 
 450,000 
 
 34 
 
 Boston, Mass., 
 
 Lake Cochituate and Sudbury River, 
 
 437,245 
 
 43 
 
 Baltimore, Md., 
 
 Lake Roland and Gunpowder River, 
 
 434,151 
 
 57 
 
 San Francisco, Cal., 
 
 Impounded water from mountain streams, 
 
 300,000 
 
 59 
 
 Cincinnati, O., 
 
 Ohio River, 
 
 296,000 
 
 67 
 
 Cleveland, O., 
 
 Lake Erie, 
 
 277,488 
 
 66 
 
 Buffalo, N. Y , 
 
 Niagara River at head, 
 
 
 
 New Orleans, La., 
 
 Drinking-water from tanks and cisterns, 
 
 254,000 
 
 20 
 
 Washington, D. C., 
 
 Potomac River, 
 
 250,000 
 
 83 
 
 Pittsburg, Pa , 
 
 Alleghany River, 
 
 . . . 
 
 
 Detroit, Mich., 
 
 Detroit River, 
 
 230,000 
 
 18 
 
 Milwaukee, Wis., 
 
 Lake Michigan, 
 
 220,000 
 
 33 
 
 Newark, N. J., * 
 
 Impounded water, Pequannock River, 
 
 181,830 
 
 60 
 
 Jersey City, N. J , 
 
 Passaic and Pequannock Rivers, 
 
 163,003 
 
 91 
 
 Louisville, Ky., 
 
 Ohio River, 
 
 161,000 
 
 88 
 
 Providence, R.I., 
 
 Pawtucket River, 
 
 132,146 
 
 29 
 
 Indianapolis, Ind., 
 
 Driven wells and Filter Gallery, 
 
 . . . 
 
 
 Lowell, Mass , 
 
 Merrimac River and driven wells, 
 
 77,696 
 
 158 
 
 Lawrence, Mass., 
 
 Filtered from Merrimac River, 
 
 44,654 
 
 123 
 
 Nashville, Tenn., 
 
 Filter gallery, Cumberland River, 
 
 77,000 
 
 46 
 
 Dayton, O., 
 
 Driven wells, 
 
 60,000 
 
 20 
 
 Covington, Ky , 
 
 Ohio River, 
 
 37,400 
 
 43 
 
 Newport, Ky.,t 
 
 Ohio River, 
 
 
 
 Denver, Col., 
 
 South Platte River and Marston Lake, 
 
 . . . 
 
 
 Atlanta, Ga., 
 
 Mechanical filter, Chattahootchie River, 
 
 65,533 
 
 151 
 
 Chattanooga, Tenn., 
 
 Mechanical filter, Tennessee River, 
 
 29,109 
 
 145 
 
 Knoxville, Tenn.,$ 
 
 Mechanical filter, Tennessee River, 
 
 40,600 
 
 101 
 
 Quincy, 111., 
 
 Mechanical filter, Mississippi River, 
 
 31,500 
 
 83 
 
 Davenport, la., 
 
 Mechanical filter, Mississippi River, 
 
 30,000 
 
 50 
 
 Montreal, Que., 
 
 St. Lawrence River, 
 
 216,300 
 
 29 
 
 Toronto, Ortt., 
 
 Lake Ontario, 
 
 167,439 
 
 93 
 
 East Jersey Water Co., Estab. April 15, 1892. 
 
 t Health Department, Estab. 1893. 
 
APPENDIX A. 
 
 269 
 
 APPENDIX A. 
 
 TYPHOID FEVER STATISTICS 
 
 OF THE PRINCIPAL CITIES OF THE UNITED STATES AND EUROPE. 
 
 Compiled from the Official Reports of Health Departments, January, 1897. 
 DEATH RATE PER 1OO.OOO OF POPULATION LIVING. 
 
 1891. 
 
 1892. 
 
 1993. 
 
 1894. 
 
 1895. 
 
 1896. 
 
 POPULA- 
 TION. 
 
 DEATH 
 BATE. 
 
 POPULA- 
 TION. 
 
 DEATH 
 BATE. 
 
 POPULA- 
 TION. 
 
 1>EATH 
 BATE. 
 
 POPULA- 
 TION. 
 
 DEATH 
 BATE. 
 
 POPULA- 
 TION. 
 
 DEATH 
 BATE. | 
 1 
 
 POPULA- 
 TION. 
 
 = _ 
 < 
 8S 
 
 1,765,645 
 
 22 
 
 1,827,396 
 
 22 
 
 1,891,306 
 
 20 
 
 1,957,452 
 
 17 
 
 1,879,195 
 
 17 
 
 1,934,077 
 
 16 
 
 1,250,000 
 
 160 
 
 1,438,010 
 
 104 
 
 1,600,000 
 
 42 
 
 1,567,727 
 
 31 
 
 1,600,000 
 
 32 
 
 1,619,226 
 
 46 
 
 1,069,264 
 
 64 
 
 1,092,168 
 
 40 
 
 1,115,562 
 
 41 
 
 1,146,000 
 
 32 
 
 1,163,864 
 
 40 
 
 1,188,793 
 
 34 
 
 880,780 
 
 20 
 
 962,530 
 
 17 
 
 990,891 
 
 17 
 
 1,045,000 
 
 15 
 
 1,090,000 
 
 16 
 
 1,140,000 
 
 15 
 
 452,000 
 
 30 
 
 460,000 
 
 37 
 
 500,000 
 
 103 
 
 540,000 
 
 31 
 
 560,000 
 
 19 
 
 570,000 
 
 19 
 
 461,093 
 
 33 
 
 474,063 
 
 29 
 
 487,397 
 
 30 
 
 501,107 
 
 28 
 
 496,920 
 
 33 
 
 508,694 
 
 32 
 
 445,853 
 
 34 
 
 458,350 
 
 42 
 
 473,193 
 
 47 
 
 455,427 
 
 49 
 
 496,315 
 
 39 
 
 507,398 
 
 37 
 
 330,000 
 
 41 
 
 330,000 
 
 34 
 
 330,000 
 
 32 
 
 330,000 
 
 35 
 
 330,000 
 
 37 
 
 330,000 
 
 31 
 
 300,000 
 
 62 
 
 305,000 
 
 40 
 
 310,000 
 
 43 
 
 336,000 
 
 50 
 
 336,000 
 
 36 
 
 341,000 
 
 48 
 
 299,475 
 
 52 
 
 309,243 
 
 54 
 
 322,932 
 
 47 
 
 325,000 
 
 27 
 
 325,000 
 
 36 
 
 330,279 
 
 43 
 
 255,664 
 
 50 
 
 285,000 
 
 34 
 
 300,000 
 
 37 
 
 315,000 
 
 36 
 
 335,709 
 
 29 
 
 350,000 
 
 20 
 
 254,000 
 
 23 
 
 254,000 
 
 21 
 
 254,000 
 
 15 
 
 275,000 
 
 28 
 
 275,000 
 
 41 
 
 275,000 
 
 33 
 
 250,000 
 
 83 
 
 260,000 
 
 70 
 
 285,000 
 
 66 
 
 270,514 
 
 71 
 
 271,000 
 
 74 
 
 278,150 
 
 51 
 
 247,000 
 
 100 
 
 255,000 
 
 100 
 
 264,000 
 
 111 
 
 272,000 
 
 56 
 
 275,000 
 
 77 
 
 280,000 
 
 61 
 
 230,000 
 
 13 
 
 230,000 
 
 51 
 
 230,000 
 
 61 
 
 250,000 
 
 26 
 
 280,000 
 
 22 
 
 279,000 
 
 20 
 
 233,333 
 
 33 
 
 245,000 
 
 31 
 
 260,000 
 
 37 
 
 267,500 
 
 26 
 
 260,000 
 
 27 
 
 257,500 
 
 18 
 
 187,108 
 
 81 
 
 192,531 
 
 45 
 
 198,115 
 
 28 
 
 203,861 
 
 15 
 
 215,725 
 
 17 
 
 230,000 
 
 21 
 
 167,237 
 
 95 
 
 171,471 
 
 53 
 
 175,000 
 
 60 
 
 179,939 
 
 76 
 
 184,173 
 
 71 
 
 187,098 
 
 61-62 
 
 161,000 
 
 81 
 
 161,000 
 
 72 
 
 161,000 
 
 84 
 
 200,000 
 
 72 
 
 205,000 
 
 77 
 
 211,100 
 
 45 
 
 132,146 
 
 47 
 
 132,146 
 
 39 
 
 148,944 
 
 34 
 
 153,000 
 
 47 
 
 145,472 
 
 32 
 
 150,000 
 
 27 
 
 120,000 
 
 36 
 
 125,000 
 
 52 
 
 125,000 
 
 106 
 
 125,000 
 
 55 
 
 125,000 
 
 97 
 
 165,000 
 
 41 
 
 80,400 
 
 98 
 
 83,200 
 
 90 
 
 87,191 
 
 61 
 
 90,613 
 
 55 
 
 84,367 
 
 39 
 
 85,700 
 
 42 
 
 45,911 
 
 115 
 
 47,204 
 
 102 
 
 48,355 
 
 93 
 
 49,900 
 
 48 
 
 52,164 
 
 31 
 
 55,000 
 
 15 
 
 80,000 
 
 56 
 
 83,000 
 
 53 
 
 85,000 
 
 24 
 
 87,000 
 
 32 
 
 87,500 
 
 47 
 
 87,754 
 
 55 
 
 60,000 
 
 32 
 
 63,000 
 
 44 
 
 75,000 
 
 64 
 
 85,000 
 
 20 
 
 80,000 
 
 47 
 
 85,000 
 
 25 
 
 40,000 
 
 45 
 
 42,500 
 
 40 
 
 45,000 
 
 27 
 
 48,000 
 
 42 
 
 48,000 
 
 27 
 
 50,000 
 
 32 
 
 
 
 120,000 
 85,000 
 
 53 
 
 87 
 
 27,500 
 125,000 
 95,000 
 
 58 
 57 
 66 
 
 30,000 
 140,000 
 108,000 
 
 37 
 35 
 43 
 
 30,000 
 145,000 
 100,000 
 
 73 
 30 
 70 
 
 30,000 
 150,000 
 110,000 
 
 63 
 61 
 60 
 
 75,000 
 
 119 
 
 34,900 
 
 06 
 
 40,000 
 
 55 
 
 36,000 
 
 86 
 
 35,751 
 
 48 
 
 40,000 
 
 47-48 
 
 40,000 
 
 30 
 
 40,385 
 
 45 
 
 40,385 
 
 37 
 
 40,385 
 
 67 
 
 40,385 
 
 59 
 
 
 
 ( 37,000 
 M 8,000 
 
 32 
 125 
 
 34,000 
 
 32 
 
 36,000 
 
 50 
 
 37,500 
 
 48 
 
 39,000 
 
 77 
 
 40,500 
 
 59 
 
 42,000 
 
 26 
 
 30,000 
 
 30 
 
 30,600 
 
 16 
 
 30,900 
 
 35 
 
 34,000 
 
 18 
 
 , 35,000 
 
 31 
 
 35,000 
 
 20 
 
 218,268 
 
 30 
 
 224,816 
 
 22 
 
 231,560 
 
 21 
 
 241,748 
 
 17 
 
 249,000 
 
 18 
 
 256,470 
 
 21 
 
 181,220 
 
 94 
 
 184,000 
 
 43 
 
 188,333 
 
 42 
 
 196,666 
 
 17 
 
 196,666 
 
 28 
 
 196,666 
 
 28-29 
 
 J (1) City proper; (2) Suburbs. 
 
270 
 
 APPENDIX A. 
 
 APPENDIX ^. Continued. 
 
 DEATH RATE PER 1OO,OOO OF POPULATION LIVING. 
 
 City. 
 
 Source of Supply. 
 
 1890. 
 
 POPULA- 
 TION. 
 
 DKATH 
 BATE. 
 
 London, Eng., 
 
 From Kent wells and filtered water from the 
 
 
 
 
 Rivers Thames and Lea, 
 
 4,180,654 
 
 16 
 
 Liverpool, Eng., 
 
 Lake Vyrnwy (Wales), 
 
 513,493 
 
 24 
 
 Manchester, Eng., 
 
 Lake Thirlmere (Cumberland), 
 
 379,437 
 
 31 
 
 Edinburgh, Scot., 
 
 Impounded water, Pentland Hills, 
 
 271,135 
 
 19 
 
 Glasgow, Scot., 
 
 Loch Katrine, 
 
 530,208 
 
 26 
 
 Dublin, Ire., 
 
 Impounded water filtered from River Vartry, 
 
 353,082 
 
 62 
 
 Paris, Fr., 
 
 Ourcq Canal, artesian wells, springs, Rivers 
 
 
 
 
 Seine, Marne, and Vanne, 
 
 2,260,945 
 
 30 
 
 Brussels (with suburbs), Bel., 
 
 
 477,288 
 
 26 
 
 Amsterdam, Hoi., 
 
 Haarlem dunes, 
 
 406,302 
 
 19 
 
 Rotterdam, Hoi., 
 
 Filtered water from River Mass, 
 
 203,486 
 
 6 
 
 The Hague, Hoi., 
 
 From sand dunes, 
 
 156,497 
 
 3 
 
 Copenhagen, Den., 
 
 Driven wells, 
 
 312,387 
 
 9 
 
 Stockholm, Sweden, 
 
 Lake and well water, 
 
 236,350 
 
 18 
 
 Christiania, Nor., 
 
 
 143,300 
 
 12 
 
 St. Petersburg, Rus., 
 
 Filtered water from River Neva, 
 
 842,000 
 
 57. 
 
 Moscow, Rus., 
 
 Mytschia springs and ponds, Moscov and 
 
 
 
 
 Yanza Rivers, 
 
 753,469 
 
 73 
 
 Berlin, Ger., 
 
 Filtered water from Lake Tegel and River 
 
 
 
 
 Spree, 
 
 1,548,279 
 
 9 
 
 Hamburg (State), Ger., 
 
 Filtered water from River Elbe, 
 
 591,647 
 
 28 
 
 Altona, Ger., 
 
 Filtered water from River Elbe, 
 
 143,249 
 
 19 
 
 Dresden, Ger., 
 
 Filter gallery by River Elbe, 
 
 269,250 
 
 9 
 
 Breslau, Ger., 
 
 Filtered water from River Oder, 
 
 324,400 
 
 15 
 
 Munich, Ger., 
 
 Spring water from Mangfall Valley, 
 
 298,000 
 
 8 
 
 Vienna (with suburbs), Aust-Hung. 
 
 Springs in the Schneeberg and driven wells, 
 
 822,176 
 
 9 
 
 Prague, Aust.-Hung., 
 
 
 314,425 
 
 33 
 
 Budapest, Aust.-Hung., 
 
 Ground water from wells, 
 
 463,017 
 
 34 
 
 Trieste, Aust.-Hung., 
 
 
 160,092 
 
 12 
 
 Rome, Italy, 
 
 Fontanadi Trevi, Aqua Felice, and Paoli, 
 
 417,392 
 
 35 
 
 Milan, Italy, 
 
 
 
 
 Turin, Italy, 
 
 
 314,827 
 
 46 
 
 Venice, Italy, 
 
 Springs in the mountains fifteen miles distant, 
 
 
 
 
 cast-iron conduit. 
 
 156,800 
 
 44 
 
 Cairo, Egypt,* 
 
 River Nile by canal, 
 
 374,838 
 
 260 
 
 Alexandria, Egypt,* 
 
 River Nile by canal, 
 
 231,396 
 
 208 
 
 Sydney (with suburbs), Austr., 
 
 Impounded water from Upper Nepean River, 
 
 
 
 Brisbane (with suburbs), Austr., 
 
 
 
 Including malarial fevers. 
 
APPENDIX A, 
 
 271 
 
 A P PE N D I X A. Continued. 
 
 DEATH RATE PER 1OO.OOO OF POPULATION LIVING. 
 
 1891. 
 
 1892. 
 
 1893 
 
 1894. 
 
 1895. 
 
 1896. 
 
 POPULA- 
 TION. 
 
 DEATH 
 KATE. 
 
 POPULA- 
 TION. 
 
 DEATH 
 RATE. 
 
 POPULA- 
 TION. 
 
 DEATH 
 RATE. 
 
 POPULA- 
 TION. 
 
 DEATH 
 RATE. 
 
 POPULA- 
 TION. 
 
 % 
 
 POPULA- 
 TION. 
 
 IS 
 
 4,222,157 
 
 15 
 
 4,264,076 
 
 11 
 
 4,306,411 
 
 16 
 
 4,349,166 
 
 15 
 
 4,392,346 
 
 14 
 
 4,421,955 
 
 14 
 
 517,116 
 
 25 
 
 513,790 
 
 25 
 
 510,514 
 
 53 
 
 507,230 
 
 58 
 
 503,967 
 
 37 
 
 632,512 
 
 32 
 
 506,469 
 
 39 
 
 510,998 
 
 25 
 
 515,598 
 
 25 
 
 520,211 
 
 18 
 
 524,865 
 
 19 
 
 529,561 
 
 23 
 
 261,970 
 
 18 
 
 264,787 
 
 13 
 
 267,261 
 
 14 
 
 270,588 
 
 15 
 
 273,535 
 
 20 
 
 276,514 
 
 16 
 
 567,143 
 
 31 
 
 669,059 
 
 18 
 
 677,883 
 
 20 
 
 686,820 
 
 24 
 
 695,876 
 
 19 
 
 705,052 
 
 23 
 
 347,312 
 
 58 
 
 349,594 
 
 39 
 
 349,594 
 
 87 
 
 349,594 
 
 48 
 
 349,594 
 
 27 
 
 349,594 
 
 45 
 
 2,424,705 
 
 20 
 
 2,424,705 
 
 28 
 
 2,424,705 
 
 25 
 
 2,424,705 
 
 29 
 
 2,424,705 
 
 11 
 
 2,511,629 
 
 11 
 
 465,517 
 
 41 
 
 476,862 
 
 23 
 
 488,188 
 
 27 
 
 498,400 
 
 14 
 
 507,985 
 
 16 
 
 518,37 
 
 18 
 
 417,539 
 
 11 
 
 426,914 
 
 15 
 
 437,892 
 
 16 
 
 446,295 
 
 8-9 
 
 451,493 
 
 11 
 
 489,496 
 
 3 
 
 209,136 
 
 4 
 
 216,679 
 
 6 
 
 222,233 
 
 5 
 
 228,597 
 
 5 
 
 272,042 
 
 2 
 
 276,338 
 
 12 
 
 . 160,531 
 
 12 
 
 165,560 
 
 4 
 
 169,828 
 
 2 
 
 174,790 
 
 3 
 
 180,455 
 
 5 
 
 187,545 
 
 4 
 
 320,000 
 
 8 
 
 330,000 
 
 7 
 
 337,500 
 
 9 
 
 341,000 
 
 7 
 
 333,714 
 
 16 
 
 333,714 
 
 7 
 
 245,317 
 
 18 
 
 248,051 
 
 19 
 
 249,246 
 
 8 
 
 252,937 
 
 8 
 
 259,304 
 
 9 
 
 267,100 
 
 6 
 
 151,130 
 
 9 
 
 156,535 
 
 4 
 
 161,151 
 
 6 
 
 167,588 
 
 3 
 
 174,717 
 
 7 
 
 182,856 
 
 33 
 
 
 
 
 
 
 954,400 
 
 51 
 
 954,400 
 
 49 
 
 954,400 
 
 87 
 
 954,400 
 
 142 
 
 753,469 
 
 75 
 
 753,469 
 
 68 
 
 753,469 
 
 40 
 
 753,469 
 
 29 
 
 753,469 
 
 50 
 
 753,469 
 
 46 
 
 1,601,327 
 
 10 
 
 1,662,237 
 
 8 
 
 1,714,938 
 
 9 
 
 1,701,643 
 
 4 
 
 1,734,492 
 
 5 
 
 1,695,313 
 
 5 
 
 622,530 
 
 23 
 
 637,686 
 
 34 
 
 634,878 
 
 18 
 
 598,372 
 
 6 
 
 608,710 
 
 9 
 
 625,552 
 
 6 
 
 144,388 
 
 64 
 
 145,527 
 
 43 
 
 146,667 
 
 15 
 
 147,807 
 
 7 
 
 148,934 
 
 13 
 
 . . . 
 
 
 276,523 
 
 8 
 
 301,400 
 
 5 
 
 308,930 
 
 4-5 
 
 316,600 
 
 8 
 
 324,341 
 
 5 
 
 342,340 
 
 4 
 
 339,000 
 
 12 
 
 346,442 
 
 15 
 
 353,551 
 
 10 
 
 360,660 
 
 6 
 
 367,769 
 
 9 
 
 377,062 
 
 8 
 
 357,000 
 
 7 
 
 373,000 
 
 3 
 
 385,000 
 
 15 
 
 393,000 
 
 2-3 
 
 396,000 
 
 3 
 
 406,000 
 
 3 
 
 1,378,530 
 
 6 
 
 1,406,933 
 
 8 
 
 1,435,931 
 
 7 
 
 1,465,537 
 
 5 
 
 1,495,764 
 
 6 
 
 1,526,623 
 
 5 
 
 310,485 
 
 37 
 
 321,167 
 
 53 
 
 327,953 
 
 36 
 
 339,172 
 
 57 
 
 351,478 
 
 46 
 
 364,632 
 
 28 
 
 513,010 
 
 23 
 
 526,263 
 
 26 
 
 539,516 
 
 15 
 
 552,769 
 
 14 
 
 566,022 
 
 20 
 
 579,275 
 
 29 
 
 156,190 
 
 11 
 
 157,343 
 
 26 
 
 158,314 
 
 17 
 
 159,739 
 
 19 
 
 160,825 
 
 5 
 
 161,886 
 
 13 
 
 427,684 
 
 36 
 
 438,123 
 
 26 
 
 449,430 
 
 34 
 
 456,777 
 
 30 
 
 465,563 
 
 62 
 
 473,296 
 
 27 
 
 
 
 424,887 
 
 62 
 
 430,829 
 
 62 
 
 . . . 
 
 
 441,948 
 
 55 
 
 
 
 320,808 
 
 41 
 
 329,724 
 
 44 
 
 334,090 
 
 29 
 
 335,957 
 
 24 
 
 344,203 
 
 32 
 
 344,203 
 
 24 
 
 158,288 
 
 33 
 
 162,664 
 
 30 
 
 163,601 
 
 26 
 
 158,187 
 
 18 
 
 158,159 
 
 23 
 
 163,254 
 
 27 
 
 374,838 
 
 235 
 
 374,838 
 
 163 
 
 374,838 
 
 154 
 
 374,838 
 
 135 
 
 374,838 
 
 90 
 
 374,838 
 
 141 
 
 231,396 
 
 348 
 
 231,396 
 406,480 
 
 77 
 20 
 
 231,396 
 411,710 
 93,657 
 
 79 
 19 
 19 
 
 231,396 
 421,030 
 93,657 
 
 100 
 29 
 10 
 
 231,396 
 423,600 
 93,657 
 
 103 
 20 
 60 
 
 231,396 
 
 89 
 
 
 
APPENDIX B.* 
 
 THE BACTERIA. 
 
 THE bacteria are minute vegetable organisms, devoid of chlorophyl, 
 and consist of a cellulose envelope containing a protoplasm described 
 as mycoprotein. 
 
 According to Nencki t they have the following chemical composi- 
 tion : 
 
 Water, 84.26 per cent. 
 
 Solids, 15.74 " " 
 
 100.00 per cent. 
 
 Of the solids Nencki finds for the putrefactive bacteria of: 
 
 Albumen, 87.46 per cent. 
 
 Fat, 6.41 " " 
 
 Ash, 3.04 " 
 
 Undetermined substances, 3.09 " " 
 
 100.00 per cent. 
 
 " The albuminous substance is not precipitated by alcohol, and 
 differs in its chemical composition from other known substances of its 
 class." 
 
 Nencki calls this substance mycoprotein, and gives the following as 
 its chemical composition : 
 
 Carbon, 52.32 
 
 Hydrogen, 7.55 
 
 Nitrogen, 14.75 
 
 * This Appendix is written with reference solely to water purification, and is intended 
 only as a brief discussion of the bacteria. Those who may desire to pursue the inquiry further 
 are referred to the standard text-books on this subject, of which may be mentioned, A "Manual 
 of Bacteriology, by Dr. George M. Sternberg, New York, 1893; The Principles of Bacteriology, 
 by Dr. A. C. Abbott, Philadelphia, 1894 ; Micro Organisms tn Water, by P. F. & G. C. Frank- 
 land, London, 1894 ; Bacteriological Diagnosis, by Dr. James Eisenberg, Philadelphia, 1892 ; 
 The Pathogenic Bacteria, by Dr. Joseph McFarland, Philadelphia, 1896, etc. 
 
 t A Manual of Bacteriology, by Dr. George M. Sternberg, New York, 1893, p. 117. 
 
 272 
 
APPENDIX B. 273 
 
 % 
 Mycoprotein contains neither phosphorus nor sulphur. 
 
 The nitrogenous body appears to vary in different species, for in 
 b. anthracis a substance has been obtained by Nencki which does not 
 give the reactions of mycoprotein. This substance he calls anthrax- 
 protein.* 
 
 The green coloring-matter of plants is known as chlorophyl, and the 
 absence of this substance in the bacteria compels them to obtain the 
 materials upon which they subsist from organic matter in process of 
 digestion or decomposition ; in fact, the destruction and splitting up 
 of organic matter into its constituent elements, is chiefly, and in some 
 cases wholly, due to bacterial agencies. 
 
 The production of carbon dioxide, carbon monoxide, and nitrous and 
 nitric acids from decomposing organic matter, is due to the action of 
 the bacteria. The putrefactive bacteria are the first to attack organic 
 matter, producing what is known as decay, with a liberation of carbonic 
 acid and other gases, while the nitrifiers discovered by Winogradsky in 
 the soil at Zurich, act upon the nitrogenous matters, and convert them 
 into nitrous and nitric acids, which, uniting with lime, sodium, potash, 
 or other bases, form the nitrites and nitrates for the support of plant 
 life. 
 
 SAPROPHYTES AND PARASITES. 
 
 The bacteria divide into two great classes : 
 
 1. Those which live and propagate their kind only upon dead or- 
 ganic matter, and known as the saprophytes. 
 
 2. Those which will live and develop only in the tissues or fluids of 
 the living body, and known as the parasites. 
 
 The line of division between the two classes is not well marked. 
 Some of the saprophytes may, under certain conditions, flourish as para- 
 sites ; while certain of the parasites, known to attain their highest state 
 of development in the animal body, will live for a limited time as sapro- 
 phytes. Thus the bacillus of tuberculosis (consumption) is classed as a 
 true parasite, but it can be cultivated (on artificial media) outside the 
 living host ; while some of the so-called saprophytes may independently, 
 or in conjunction with certain of the pathogenic bacteria, be responsible 
 for processes in the animal body which result in disease, and should 
 therefore be regarded as facultative parasites. 
 
 Bacteria which cannot subsist upon living matter are strict sapro- 
 phytes, while those which cannot subsist upon dead matter are true 
 parasites ; but the dividing line i.j not so distinct that we can readily 
 
 * A Manual of Bacteriology, by E. M. Crookshank, London, 1890, p. 148. 
 
274 APPENDIX B. 
 
 determine with regard to certain bacteria whether they are the one or 
 the other ; and a saprophyte may be &. facultative parasite, while a parasite 
 may be a facultative saprophyte. 
 
 LIQUEFIERS AND NON-LIQUEFIERS. 
 
 The bacteria again divide into two other great classes : 
 
 1. Those which when cultivated in gelatin will render it fluid, and 
 known as the liquefiers. 
 
 2. Those which will develop on or in gelatin without liquefaction, 
 and known as the uon-liquefiers. 
 
 The bacillus of typhoid fever will not liquefy gelatin, while the bacil- 
 lus of cholera does liquefy gelatin. Most of the pathogenic or disease- 
 producing bacteria are uon-liquefiers, while most of the putrefactive bac- 
 teria are rapid liquefiers. Certain of the bacteria will liquefy gelatin at 
 room temperature (70 Fahr.) within a day or two, while others require 
 a growth of two or three weeks to render the gelatin fluid, and some 
 reduce the solid gelatin to a fluid at a rate so slow that the water of 
 liquefaction is evaporated through the cotton plug of the test-tube as 
 rapidly as it is formed. 
 
 The liquefaction of gelatin by bacterial agencies is not due to the 
 production of heat in the destruction of organic matter, but to certain 
 somewhat indistinct changes, by which the gelatin is peptonized and 
 rendered incapable of again becoming hard at the temperature of melt- 
 ing ice (32 Fahr.). 
 
 AEROBIANS AND ANAEROBIANS. 
 
 Again, the bacteria divide into two other great classes as defined by 
 Pasteur : 
 
 1. Those which will grow only in the presence of oxygen, and 
 termed aerobians. 
 
 2. Those which will not grow in the presence of oxygen, and termed 
 anae'robians. 
 
 Most of the bacteria, so far as we are now aware, are aerobians ; 
 but it is probable that plate cultivation under strictly anaerobic condi- 
 tions may demonstrate the existence of species now unsuspected which 
 will not grow in the presence of oxygen. For example, the bacteria 
 found upon cultivation of a sample of water in a Petri dish, or on a 
 plate, are all aerobians, while the bacillus of tetanus (lockjaw) is a true 
 anaerobian. Under the usual conditions of plate culture the anaerobic 
 species, if any are in the water or other sample, will not grow, and of 
 
APPENDIX B. 275 
 
 course do not enter into the subsequent count of the colonies, nor into 
 the differentiation of species found on the plate. 
 
 The plugging of the glass test-tubes for culture media with cotton, 
 allows the air to pass freely into the tube, while it effectually prevents 
 the entrance of any organism in the air, however small it may be. 
 
 Naturally, because of the difficulties of cultivation, few anaerobic 
 species of the bacteria have been found in water, but with improved 
 and more convenient methods of (anaerobic) cultivation, more may in 
 the future be found. 
 
 FORMS OF THE BACTERIA. 
 
 The bacteria are seen to be of three general forms when examined 
 under the microscope. 
 
 1. The Cocci, or spherical forms, which for convenience of identifica- 
 tion are divided into : 
 
 The Micrococci, when the spheres are single or in irregular groups. 
 
 The Diplococci, consisting always or occasionally of two spheres 
 joined, and resembling a dumb-bell with the connecting rod missing. 
 
 The Tetrads, which consist of triangular groups of three of the 
 spheres, or cocci. 
 
 The Streptococci, in which the spheres are found in chains of many 
 members. 
 
 The Staphylococci, in which the spheres, or cocci, are grouped some- 
 what like bunches of grapes. (To this class belong the yellow and 
 white germs of septicaemia, or blood poisoning.) 
 
 The Sarciiia, in which the spheres are found in cubical packets, 
 divided in three directions, like a bale of goods tied with cords parallel 
 to all the sides. 
 
 The sarcina contain 8 or more spheres. Thus, if divided once in 
 each direction, the packet will contain the cube of two, or 8 spheres ; if 
 divided twice in each direction, the packet will contain the cube of 
 three, or 27 spheres. Usually the sarcina are seen under the microscope 
 as broken packets, the preparation and fixing of the culture on the 
 cover glass breaking up the characteristic arrangement of the members. 
 
 The cocci are regarded as non-spore-bearing bacteria. 
 
 2. The bacilli, or rods, straight or slightly bent. 
 
 To this class belong the organisms b. typhosus, diphtheria, c'oli corn- 
 munis, lactis aerogenes, the tubercle bacillus, and nearly all the putrefac- 
 tive and pathogenic bacteria. 
 
 The bacilli grow into rods, and separate into individual cells by 
 
276 APPENDIX B. 
 
 fission. Thus one rod becomes two rods by separation in the middle ; 
 each of these separates again ; and as an evidence of the rate of growth 
 and multiplication of some of the bacteria, it is stated that one rod or 
 cell may, under favorable conditions, grow and divide within twenty 
 minutes ; from which, by calculation, it will be seen that one rod may 
 become the parent of nearly 17,000,000 within twenty-four hours.* 
 
 Of the bacilli, most are straight rods with round ends. B. anthratis 
 is a straight rod with square ends united in chains. In young cultures 
 of the typhoid germ long crooked strings are frequently noticed ; these 
 strings, or filaments, consist of many bacilli united together. In due 
 time such strings, by fission, separate into the typical bacilli of varying 
 length. The manner in which the bacilli are displayed on a cover glass 
 preparation is an important number in the differentiation of species. 
 Thus certain kinds, like b. anthracis, always are found in well-defined long 
 chains, with some detached links or rods (probably broken from the chain 
 by the manipulation of the specimen on the cover glass) ; others occur 
 only as separate rods ; sometimes short chains, of two to six or eight rods, 
 constitute the manner of grouping, but in all cases there is a method of 
 grouping which is a property or characteristic of the species. 
 
 3. The spirilla, or vibrios, are rods always bent, sometimes in the 
 form of the letter " S." To this class belongs the comma bacillus of Koch, 
 known as the cholera germ. As a rule, the spirilla are rods which, 
 if measured on the curve, are longer than the bacilli ; or the bacilli 
 and spirilla may both be regarded as rod forms of the bacteria, the 
 bacilli usually being the shorter and straight rods, while the spirilla are 
 the longer and always bent or crooked rods. The germ of diphtheria is 
 a comparatively long, slightly bent rod, but is classed as a bacillus. 
 
 The spirilla found in water are much fewer in number of kinds than 
 the bacilli ; in fact, the majority of the bacteria found in polluted waters 
 are of the latter form. 
 
 MOTILITY OF THE BACTERIA. 
 
 A property of the bacilli and spirilla, the cause of which is still open 
 to investigation, is motility. Certain of the bacteria of these forms 
 when examined, stained or unstained in drop cultures, exhibit surprising 
 activity. Thus the germ b. typhosus (in a drop of bouillon) has motions 
 of translation and rotation, and sinuous movements like a snake. Occa- 
 sionally some of the rods, when taken from young cultures, will be 
 
 * A Manual of Bacteriology, by Dr. George M. Sternberg, p. 114. 
 
APPENDIX B. 277 
 
 observed in rotation resembling the movement of am acrobat on a hori- 
 zontal turning-bar, while others have a sluggish or no motion at all. 
 B. fluorescens Uquefaricns, a bacterium frequently found in water, re- 
 sembles in motility b. typhosus ; it resembles also the smaller rods of 
 this germ in dimensions ; but unlike b. typkosus, it rapidly liquefies the 
 gelatin in which it is grown, and in gelatin, and especially upon agar, 
 produces a beautiful fluorescent green which permeates the whole cul- 
 ture material. 
 
 B. coli communis has motility, but the motions of the bacillus are 
 sluggish and unlike those of b. typhosus. B. lactis aerogenes is not pos- 
 sessed of motility. 
 
 Motility of the bacteria, when viewed in drop cultures, is not to be 
 confounded with pedesis, which is a swaying or oscillatory motion of the 
 organism not due to inherent powers of locomotion. The motile bacilli 
 and spirilla are provided with delicate flagella or hair-like appendages, 
 which, acting as whips or oars, propel the germ through the drop of 
 fluid on the cover glass. It was at one time supposed that the motility 
 of the bacilli bore some relation to the number of the flagella, but re- 
 cent investigations seem to negative this belief.* 
 
 In Germany the flagella have been regarded as an important ele- 
 ment in the differentiation of the bacteria, but according to Dr. V. A. 
 Moore, the flagella cannot be taken into serious consideration in the 
 differentiation of closely allied species. f 
 
 With reference to the form, size, and other features of the bacteria, 
 due allowance must be made for the environment of the culture. The 
 culture media, its reaction, temperature of incubator, and nearness of 
 the culture to its original source, all have an important bearing on the 
 differentiation of species. When the bacteria are grown in a Petri dish, 
 room temperature (about 70 Fahr.) is preferable as corresponding with 
 the conditions under which the largest and most rapid growth will be 
 obtained; but at room temperature most of the pathogenic bacteria de- 
 velop slowly, and even if such were in a water sample, the probability of 
 finding them in plate cultures is rather remote. 
 
 CHROMOGENIC SPECIES. 
 
 Certain of the bacteria when grown upon suitable materials elab- 
 orate beautiful colors, which rival the colors of the solar spectrum. 
 For example, b. prodigiosus, a bacillus found in water, produces a deep 
 blood red ; b. rothe, another water bacillus, produces a raspberry red ; 
 
 * Character of the Flagella, V. A. Moore, Washington, D.C., 1893. t Ibid., P- 363. 
 
278 APPENDIX B. 
 
 b. violaceus, also a water bacillus, produces a purple merging into blue ; 
 the staphylococcus pyogenes aureus, the organism of malignant pustule or 
 septicaemia, found in water, produces a golden yellow ; b. proteus fluo- 
 rescens produces a fluorescent green ; m. candicans, a water germ, pro- 
 duces a dazzling Chinese white ; m. aurantiacus, another water germ, 
 elaborates a beautiful orange color ; and m. carneus, a water micrococci, 
 produces a delicate pink or flesh color when grown on agar. The color, 
 when a characteristic, is an important element in determining species. 
 
 Among the products of bacterial action on dead organic matter are 
 the ptomains, some of which have toxic properties. The substance iso- 
 lated by Dr. V. C. Vaughan * from ice-cream and cheese, called tyrotoxi- 
 con, is one of the vital products of the putrefactive bacteria. Whether 
 the putrefactive bacteria are capable of producing ptomains from the 
 organic matter in water is not known, but some of the investigators 
 abroad seem to suspect the possibility of it. 
 
 The action of the pathogenic bacteria on organic matter is the pro- 
 duction of toxins, which probably are absorbed into the circulation of 
 the animal with the symptoms during life characteristic of specific dis- 
 ease, and the pathological lesions usually found upon post-mortem ex- 
 amination. The toxin from the growth of the bacillus of diphtheria on 
 the mucous membrane of the fauces, when taken into the circulation, 
 produces the symptoms and lesions characteristic of this disease. Dr. 
 McFarland remarks upon the virulent properties of the toxin elaborated 
 by the diphtheria bacillus,^ " No more convincing proof of the existence 
 of a powerful poison in diphtheria could be desired than the evidences 
 of general toxaemia, resulting from the absorption of material from a 
 comparatively small number of bacilli situated upon a little patch of 
 mucous membrane." 
 
 DIMENSIONS OF THE BACTERIA. 
 
 The dimensions of the bacteria are stated in microns, designated 
 by the Greek letter " //,," which is T <jW f a millimeter, equal to about 
 5 _i^_ of an inch. Thus the typical dimensions of b. typhosus are .5 to .8 
 "/A" wide, by 1.5 to 2.5 "/A" long; or about s^oo to 57550 of an mc ^ 
 wide or thick, and TF ^ ff to TTT ^<j of an inch long. 
 
 Taking the average length of the typhoid bacillus as 2 microns (/x), 
 it will be seen that it would require 12,500 of these little rods placed 
 end to end to make one inch. The human mind can scarcely grasp the 
 
 * The Ptomaines and Leucomaines. by Vaughan and Novy, Philadelphia, 1891, p. 35 et seq. 
 | A Treatise on the Pathogenic Bacteria, Dr. Joseph McFarland, Philadelphia, 1896, p. 227. 
 
APPENDIX B. 279 
 
 smallness of the bacteria ; but assuming that the un&ided vision is capa- 
 ble of distinguishing 200 lines or divisions to the inch, then each of 
 such divisions would contain over 60 of the typhoid germs, placed end 
 to end, or 180 if placed side by side. 
 
 The air and soil both contain bacteria which may come into water, 
 and aside from the pathogenic species, all bacteria found in water must 
 not be regarded as indigenous to this source. The natural water bacte- 
 ria may be considered as those found in the water from deep wells after 
 the rainfall has percolated through many feet of various kinds of soil 
 and filtering material, and even these may be, and probably are, from 
 extraneous sources; but for the present purpose it may be held that 
 such belong to water because of the inability of the filtering materials 
 in the drift and rock to restrain them. 
 
 Rain is probably free from bacteria and organic matter as it falls 
 from the clouds ; but in falling, material suspended in the atmosphere 
 will be intercepted and carried down to the earth, and into the usual 
 receptacles or channels of discharge of rainfall. In addition to the 
 bacteria and organic matter from the air, bacteria and matter from the 
 soil is washed into the streams and lakes ; and the excess in numbers 
 and gain in species of bacteria in river or lake waters, over those in 
 deep well waters, may be attributed to the air and soil, or to sewage 
 pollution. Any bacteria naturally in the air will be intercepted by rain- 
 fall ; and if the species are capable of an independent existence in 
 water, may upon examination be found there. Likewise storm water 
 discharged through natural channels will contain bacteria intercepted in 
 flowing over the ground, together with some species obtained from 
 erosion of the earthy banks. 
 
 In the examination of a water sample for bacteria, a large number, 
 or rapidly liquefying organisms, should suggest the probable presence 
 of the putrefactive germs, among which are often found b. proteus vul- 
 garis, b. meseutericus vulgatus, and others of like character, which usually 
 are held to come into water from sewage sources. 
 
 SPORE-BEARING BACTERIA. 
 
 A characteristic of the bacteria not to be overlooked in differentia- 
 tion for species, is the presence or absence of spores. This cannot, in 
 all cases, be easily determined ; but a germ which yields spores is known 
 to be much more difficult to destroy than non-spore-bearing germs. 
 The spores are small, round, ovoid or oblong bodies, of which one or 
 more may be noticed in a single bacillus or spirillum, which will live 
 
280 
 
 APPENDIX B. 
 
 and propagate bacteria of its kind after the destruction of the germ 
 itself. 
 
 Of the pathogenic bacteria found in water a few develop spores. 
 " According to H ueppe, the Koch comma bacillus forms arthrospores, but 
 it possesses no form which is endowed with any considerable powers of 
 resistance.'" This view is not shared by all investigators.! B. an- 
 thracis, another pathogenic germ found in water, forms spores. The 
 tetanus bacillus forms a spore in one end of the rod which gives it the 
 form of a drumstick. B. pyocyanus, found in green pus and also in 
 water, forms spores. Bacillus of mouse septicaemia forms spores. Of the 
 twenty-three pathogenic species of bacteria found in water, the above 
 are all that are certainly known at present to form spores. 
 
 The Franklands in their work on Micro Organisms in Water, give a 
 list of 200 species of the bacteria which have been found in water, of 
 which the following are classed as pathogenic varieties : 
 
 PATHOGENIC BACTERIA FOUND IN VARIOUS WATERS. 
 
 GERM. 
 
 ORIGINAL DATE OF 
 IDENTIFICATION. 
 
 AUTHORITY. 
 
 B. anthracis, 
 
 1850 
 
 Rayer and Davaine, Pollender-Pasteur, 
 
 
 
 and Joubert-Koch. 
 
 B. typhosus, 
 
 1880 
 
 Eberth-Gaffky. 
 
 B. mouse septicamia, 
 
 1881 
 
 Gaffky-Loffier. 
 
 B. rabbit septic&mia. 
 
 1881 
 
 Koch-Gaffky. 
 
 B. pyocyanus, 
 
 1882 
 
 Gessard-Charrin-Ernst. 
 
 Sp. Asiatic cholera. 
 
 1884 
 
 Robert Koch. 
 
 B. tuberculosis, 
 
 1884 
 
 " 
 
 B. saprogenes, 
 
 1884 
 
 Rosenbach. 
 
 Staph. pyogenes aureus, 
 
 1884 
 
 Rosenbach-Passet-Fick. 
 
 B. coli communis, 
 
 1885 
 
 Escherich. 
 
 B. lactis aerogenes, 
 
 1885 
 
 " 
 
 B. proteus vulgaris, 
 
 1885 
 
 Hauser. 
 
 B. proteus tnirabilis, 
 
 1885 
 
 " 
 
 B. proteus Zenkeri, 
 
 1885 
 
 " 
 
 B. brevis, 
 
 1888 
 
 Rintaro Mori. 
 
 B. capsulatus, 
 
 1888 
 
 " 
 
 M. biskra, 
 
 1888 
 
 Heydenrich. 
 
 B. of tetanus, 
 
 1889 
 
 Nicolaer-Kitasato. 
 
 Coccus " B," 
 
 1890 
 
 Foutin. 
 
 B. hydrophilus fuscus, 1891 
 
 Sanarelli. 
 
 B. proteus fluorescent, 1892 
 
 Jaeger. 
 
 Sp. Berolin&nsus, 1893 
 
 Neisser. 
 
 B. tholoeideum, 
 
 Gessner. 
 
 * Micro Organisms in Water, Percy & Grace Frankland, London, 1894, p. 399. 
 f Principles of Bacteriology, by Dr. A. C. Abbott, Philadelphia, 1894, p. 314. 
 
APPENDIX B. 281 
 
 In addition to the pathogenic bacteria heretofore found in water, 
 certain investigators abroad seem to think that the bacillus of diphtheria 
 may be transmitted through the medium of water supply. In his evi- 
 dence before the Royal Commission on Metropolitan Water Supply, Dr. 
 Alfred Ashby * stated a belief that diphtheria might be so transmitted. 
 Dr. E. Frankland t says that animal refuse finding its way into water 
 may be accompanied by zymotic poisons dangerous to health, such as 
 those of typhoid fever, tuberculosis, or diphtheria ; while Dr.. George 
 Turner $ testified before the Commission that, " he had had one case 
 where he suspected that diphtheria was conveyed by water." No proof 
 is at hand indicating the transmission of this germ by water, although 
 it is possible that it may find its way into water, in the same way as 
 the tubercle bacillus, by the sputa or membraneous sloughings from a 
 patient suffering with this disease. It, however, may be said that the 
 remedies to be applied to polluted waters, or the precautions to be 
 observed in selecting water from the best natural sources, will have 
 the same influence in diminishing the chances of propagation of diphthe- 
 ria (if it should be shown to have a temporary habitat in water) as upon 
 the transmission of typhoid fever and cholera by this means. 
 
 In order to render the bacteria easily discernible under the micro- 
 scope, recourse is had to dilute solutions of the aniline dyes, which as 
 simple watery solutions, or in combination with a weak acid or alkali, are 
 readily taken up by the protoplasm of the cell substance. The bacteria 
 being devoid of chlorophyl, and consequently colorless in drop cultures 
 or fixed on cover glasses unstained, are somewhat difficult to study; 
 while by the addition of the dyes, or stains, these colorless bodies become 
 more or less opaque, and contrast sharply with the light transmitted 
 through the preparation, and when stained are easily viewed and exam- 
 ined microscopically. 
 
 The manner in which the bacteria take the stain is a material ele- 
 ment in the differentiation of species, and in jotting down the memo- 
 randa of examination of an organism the experienced observer never 
 fails to note the facility with which the stain is taken up. B. anthracis 
 thus takes the simple watery stain readily, while b, typhosus can be 
 promptly colored only by an acid or alkaline solution of the stains, and 
 the tubercle bacillus stains with great difficulty. The bacillus of tetanus, 
 on the other hand, is easily colored with the watery solutions, while 
 
 * Report of Royal Commission on Metropolitan Water Supply, London, 1893, Minutes 
 of Evidence, p. 140. 
 
 f Ibid., Appendices to Minutes of Evidence, p. 200. 
 | Minutes of Evidence, p. 177. 
 
282 APPENDIX B. 
 
 b. proteus fluorescens requires the strongest dyes to give it color. This 
 property of taking stains is affected by the age of the culture, old cul- 
 tures being more troublesome to stain than young ones. 
 
 To illustrate the importance of familiarity with the action of the 
 aniline dyes on the bacteria, if a cover glass preparation suspected of 
 being the typhoid bacillus was under examination, and it took the watery 
 solutions promptly, it can be safely set down that it is not b. typhosus, 
 but some other germ. 
 
 The growth of the common water bacteria is inhibited by high tem- 
 peratures, while the pathogenic germs attain their highest development 
 at the temperature of the body. Thus, while the waters in rivers, lakes 
 and reservoirs, attain the highest temperature at the end of summer, a 
 temperature unfavorable to the growth of many of the water bacteria, 
 they are approaching the condition favorable to the growth and full 
 development of the pathogenic organisms ; and it is at this season of 
 the year when water-borne diseases should be most manifest. 
 
 In the following tables are given : 
 
 1. A list of bacteria found in water which resemble the typhoid 
 bacillus in some dimension, and like b. coli communis and b. lactis aero- 
 genes, resemble it sometimes in other respects. 
 
 2. A list of the germs smaller than b. typhosus found in water. 
 
 3. A list of the larger bacteria found in water. 
 
 4. A list of the spore-bearing bacteria which have been found in 
 water. 
 
 In all the tables the principal properties of the organisms are given 
 to assist in the rough identification or differentiation of species. But 
 for the exact differentiation, the best descriptions of the various bacteria, 
 together with long and conscientious experience, will be found abso- 
 lutely necessary.* 
 
 * In the author's forthcoming work on the "Interpretation of Water Analysis," a full 
 description will be given of the modern methods of Bacterial Water Analysis, and the aids to 
 identification of species of bacteria. 
 
APPENDIX B. 
 
 283 
 
 
 U) 
 
 
 
 
 
 
 
 *j 
 
 
 
 
 
 
 c 
 
 
 O 
 
 I 
 
 
 TJ 
 
 C 
 
 
 
 
 c 
 
 t* 
 
 
 
 a! a 1 
 
 eg B JS 
 
 ARKS 
 
 : 
 
 o 
 
 B 
 
 2. 
 
 '-C 
 
 o 
 p 
 
 
 *o 
 
 s 
 
 i 
 
 
 j 
 
 S 
 m ^JQ 
 
 1 1 
 
 S E 
 |"J 
 
 
 
 li Si s*^ 
 
 2s 5 " 3 c 
 
 MO t* (fl^j 
 
 i 
 
 i 
 
 ded ends 
 nents. 
 
 ded ends 
 
 
 T3 
 C 
 rt 
 
 T3" 
 1 
 
 ded ends 
 
 
 
 'c 
 
 g I. 
 ||" 
 
 1 B 
 
 ** C 
 
 c *- 
 
 -Z 
 C 
 
 * 
 
 s chains 
 
 C ^ -o s "S 'M .C 
 
 
 11 
 
 K 
 
 s 
 
 3 
 
 e 
 
 
 
 
 C 
 3 
 O 
 
 
 c 
 
 3 
 O 
 
 PIS 
 
 I 
 
 C 
 
 y 
 
 II 
 
 o o 
 
 1 
 
 | 
 
 ii lilJP 
 
 a 
 
 
 
 
 
 
 ^ 
 
 
 
 
 
 a 
 
 z s- 
 7 
 
 
 fc 
 
 - 
 
 , 
 
 
 
 S 
 
 >> 
 
 JiT ^ 
 
 
 
 
 
 o 
 
 
 
 
 
 
 
 a 
 
 1 1 
 
 \ 
 
 
 1 -1 
 
 u, w 
 
 S! 
 
 
 
 
 
 * 
 
 
 
 X * 
 
 
 
 
 x 
 
 1 
 
 -5 
 
 
 1 
 
 
 
 
 
 
 
 
 E- 
 
 >> tn* 
 
 >, 
 
 
 ^, 
 
 
 
 
 >> 
 
 
 
 > >* 
 
 
 
 
 
 
 
 
 
 
 
 
 
 w 
 
 -0 Wl 
 
 IS 
 
 
 
 3 
 
 
 
 
 
 TJ 
 
 
 
 XI "73 
 
 
 
 3 
 
 
 i 
 
 
 Sb 
 
 
 
 
 
 
 
 n 
 
 H 
 
 H 
 
 
 H 
 
 
 
 
 h 
 
 
 
 H H 
 
 z 
 
 T 
 
 
 
 
 
 p 
 
 g. 
 
 
 
 
 - 
 
 H 2 
 
 S 
 
 
 
 
 
 ^^ 
 
 ^ . 
 
 
 
 
 2 . 
 
 
 1 
 
 " 
 
 ~* - 
 
 - " 
 
 - 
 
 " 
 
 * 
 
 * 
 
 
 
 Q. 
 
 CO pj 
 
 o 
 
 
 
 
 
 
 
 
 
 
 in 
 
 , ^ ^ 
 
 2 
 
 K 
 
 
 
 
 
 
 
 
 
 
 
 , * 
 
 (T 
 
 
 
 
 
 
 
 r .2" 
 
 
 
 S 
 
 & H 
 
 
 c 
 
 
 
 s 
 
 s : 
 
 = = 
 
 s 
 
 5 
 
 5 5 
 
 0" _ T 
 
 M " o 
 
 S 
 
 5 
 
 3 S jj 5 5 3 is 
 
 ^ 5 
 
 fe 
 
 
 
 
 
 
 
 
 
 
 
 
 3 
 
 
 
 
 
 
 
 p 
 
 J8J? 
 
 JU __ 
 
 
 g 
 
 r.F.5f ~ jf jif 
 
 |j 
 
 | 
 
 o 
 g 
 
 Z 
 E 
 
 .a 
 
 | 
 
 o 
 E 
 
 o o 
 S 
 
 I la" 
 
 | 
 
 c 
 
 B 
 
 inn ii 
 
 H 
 
 >> 
 
 
 >^ 
 
 
 Q 
 
 . 
 
 >> >< 
 
 >* >* O 
 
 o 
 
 
 
 g 
 
 
 
 3 
 
 > s 
 
 
 6 
 
 7: 
 
 SJ SJ 
 
 gj-J S 
 
 E 
 
 B 
 
 *g 1> 1> ~j> 
 
 H 
 
 3 
 
 .f 
 
 '5 o 
 
 fe 
 
 o 
 
 bfl 
 
 U 
 
 ~S 
 
 5 
 
 .tf 
 
 .? 'S *o -^ ' o '^ .SP 
 
 
 <! 
 
 en 
 
 < 
 
 ^ ^ 
 
 * 
 
 CO 
 
 <! < 
 
 <! < 
 
 ^ 
 
 S 
 
 s<<^^a-<S 
 
 00 
 
 !l 
 
 4 
 03 
 
 S 
 
 41 
 
 i* 
 
 
 1 
 
 a. jj 
 
 10 c 
 
 IH * ^ 
 
 
 
 M 
 
 J" -* * J" 
 
 7 c ?* e fg4 
 
 s 
 
 o 
 
 1 
 
 bo bo 
 
 S i 
 
 
 f^ 
 
 ^ "a^ 
 
 
 '-i 
 
 Lj . 
 
 !3!3!2r-5rt!2 c l " 
 
 B 
 
 X 
 
 X 
 
 In IS " 
 
 ^ x 
 
 
 X 
 
 x a > 
 
 XXX 
 
 
 x 
 
 
 s 
 
 1 
 
 ^ 
 
 S E| 
 
 4 
 
 tf 
 
 CO 
 
 ^ v a 
 
 10 oo t-. 
 
 1 
 
 [: 
 
 t_ 's S 
 
 
 := 1 
 
 
 cS ** 
 
 n4 
 
 
 
 ' 3~ 
 
 q 
 
 55 
 
 
 * 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 c 
 
 
 
 
 
 
 
 
 
 
 
 
 u 
 
 
 
 
 
 
 
 iH (N 
 
 1C 
 
 
 
 wT 
 
 o 
 
 
 
 " w" *" 
 
 
 
 
 
 
 
 
 
 
 
 
 
 GEBM. 
 
 . Typhosus, 
 
 Coli communis, 
 
 Aquatilis sulcatus 
 Aquatilis sulcatus 
 
 Aquatilis sulcatus 
 Lactis aerogenes, 
 
 Tholoeideum, 
 
 Fluorescens tenui 
 
 Fluorescens longi 
 Viridis pallescens 
 
 Fluorescens lique 
 Termo, 
 Rabbit septicaemi 
 
 r eisser bacillus 
 
 w" 
 
 ! !! 
 
 C 5 g" w - c 
 
 
 CQ 
 
 CQ 
 
 CQ PQ 
 
 ffl PQ 
 
 CQ 
 
 PQ 
 
 CQ CQ 
 
 CQ CQ CQ 
 
 ^ 
 
 CQ 
 
 fflMfflfflPQCQpqffl 
 
284 
 
 APPENDIX B. 
 
 
 .S 
 
 
 
 
 
 
 
 ( - 
 
 
 
 
 
 JS 
 "w 
 
 
 
 
 
 
 
 S 
 
 
 
 2 
 
 
 M 
 
 
 
 
 
 
 
 i 
 
 
 
 
 
 
 c 
 
 & 
 
 
 
 
 
 
 
 !2 
 
 
 o 
 
 M 
 
 K 
 
 "(3 
 A 
 
 m 
 a 
 
 
 
 Ii 
 
 -n 
 
 
 
 c 
 
 
 
 
 If 
 
 ng wavy 1 
 
 aments. 
 d in chair 
 
 i pairs, 
 orners. 
 
 aments. 
 nents in j 
 
 
 
 
 C T3 
 
 - * 
 
 
 -a 
 
 8 
 
 C T3 
 
 c 
 
 < - 
 
 
 S S3 
 
 
 o 
 cr 
 
 al 
 
 
 
 c 
 
 
 ? ^ 
 
 I 
 
 G ^ 
 
 1 
 
 c ,,_, 
 c 
 
 
 -1 
 
 w 
 
 
 
 Pi 
 
 
 O C 
 
 ta 
 
 fe W 
 
 tt 
 
 fe n 
 
 oS 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 g 
 
 
 1 
 
 
 
 
 
 
 
 
 ^ 
 
 
 w 
 
 < ? 
 
 'O 
 
 ' 
 
 
 
 
 
 
 
 
 | 
 
 
 
 
 s 
 
 o 
 
 1 
 
 
 
 
 
 
 
 1 
 
 
 , 
 
 
 
 
 
 
 
 
 
 
 
 
 ii 
 
 
 
 
 
 
 
 
 
 
 
 
 is 
 
 
 
 
 
 
 
 
 
 
 
 
 s * 
 
 
 
 
 
 
 
 
 
 
 
 
 z 
 
 o 
 
 g 
 
 
 | 
 
 
 
 
 
 
 s 
 
 
 
 
 
 
 . 
 
 
 
 
 2 . 
 
 S 
 
 1 
 
 
 
 E s 
 
 cc a 
 
 
 p 
 
 
 VI 
 
 
 
 
 
 
 o. . 
 
 in 
 
 1 
 
 
 
 
 
 
 
 
 O 
 
 
 
 
 
 o 
 
 o 
 
 
 
 
 fc 
 
 
 
 
 
 
 ^ 
 
 ^ 
 
 
 
 
 , g 
 
 
 J 
 
 -.2" 
 
 
 
 
 .2- - 
 
 
 
 | 
 
 
 63 S 
 
 .2* 
 
 s c 
 
 .2" c 
 
 
 
 
 
 .5- 
 
 c .2" 
 
 - 
 
 - 
 
 c : 
 
 cr _ 
 
 
 
 
 
 
 
 
 ; 
 
 
 
 
 
 
 rf * 
 
 
 
 
 ' 
 
 
 
 
 
 
 
 
 fc 
 
 
 
 
 tT 
 
 T 
 
 
 
 
 
 
 
 
 
 
 
 T ,7 
 
 
 
 2 
 
 ll 
 
 ~ 
 
 
 
 c 
 
 S 
 
 | 
 
 
 
 
 E 
 
 ' 
 
 11 
 
 i 
 
 
 >- o 
 "i; E 
 
 >> o 
 
 
 
 X 
 
 H 
 
 i " 
 
 - 
 
 >> 
 
 1 "^ 
 
 x 
 
 o 
 a 
 
 c 
 o 
 
 
 
 
 
 t! 
 
 c 
 
 II 
 
 1 
 
 ' 
 
 H 
 
 55 < 
 
 > 2 
 
 < c/5 
 
 
 
 
 
 
 4 
 
 
 
 
 i 
 
 
 
 
 ^ . 
 
 . *: 
 
 4 . 
 
 . 
 
 'ii 
 
 
 
 4 
 
 
 OT 
 
 4 
 
 
 H 
 e 
 
 < 
 
 *3 
 
 oi> ^ 
 
 sl 
 
 * 
 
 4 
 
 1 
 
 =3? 
 
 4 
 
 rt 
 
 ^l 
 
 ^ 4 
 
 2 2 
 
 4 4 
 TH eo 
 
 2 
 
 X 
 
 x . 
 
 X 
 
 
 X 
 
 
 i X 
 
 X 
 
 X (/) 
 
 X X 
 
 X X 
 
 S 
 
 5 
 
 4 
 
 1 ^| ! 
 
 ? ^ 
 
 |l 
 
 !! 
 
 I 
 
 4 
 . 
 
 H 
 
 1* 
 
 4 
 
 4| 
 
 4 4 
 
 1 2 
 
 4 
 
 
 
 S 
 
 
 r 
 
 
 
 
 
 oj 
 
 
 (U 
 
 
 
 
 
 
 
 
 
 
 & 
 
 
 GO 
 
 
 
 icfaciens. 
 
 en 
 
 
 
 
 
 
 
 
 
 
 
 
 C 
 
 
 
 
 
 
 
 - 
 
 
 
 
 
 
 8 
 
 
 
 
 
 
 
 y 
 
 
 
 
 
 M 
 
 
 
 
 
 
 
 G 
 
 
 
 
 
 
 
 
 
 
 M 
 
 
 
 V 
 
 
 1 
 
 o 
 
 Mouse sept., 
 
 Pyocyanus, 
 Fluorescens n( 
 
 Tetanus, 
 Aquatilis fluor 
 
 Saprogenes, 
 
 Rubefaciens, 
 
 llow bacillus, 
 
 Flavocoriaceu 
 Tremelloides, 
 
 Nubilis, 
 
 Cuticularis, 
 icillus "C," 
 
 Acidi lactici, 
 Lactis cyanog 
 
 Butyricus, 
 Aquatilis grav 
 
 
 n 
 
 - CQ 
 
 CQCQ 
 
 CQ 
 
 ffl 
 
 
 CQ m 
 
 DQ 
 
 CQ m 
 
 n CQ 
 
 CQ PQ 
 
APPENDIX B. 
 
 285 
 
 
 J2 "S 1 - J9 
 
 a ti -S < T-I 2 
 
 " 
 
 
 11 i* 1 
 
 Sl '5 ^ | S 
 
 if 
 
 
 ^ JS - T3 
 
 . X =*- jo 1- 
 
 
 REMARKS. 
 
 as s! 1 
 
 'S ^ > S 
 
 13" 1 ?8 a" 3* j$ * 
 
 S 3 g g g S ! *g g 
 
 rtf ,^3 > ojrtnJ c 3 r d*-' r Oc3 
 bfl bJO^ S.Ssb/JbJQ'SSS^t^ 
 hJ J W :/: M M eft P* &EI & Cfi 
 
 iiift || 1 
 
 ,1! ll!!:lll 1 
 'ill tl|l!i|lf| 
 
 11 9 -Slllilpll 
 - ,2^ II*IS E ll T 4 
 
 into ammonium carbc 
 Spores are meniscus, s 
 or club-shaped. 
 
 j 
 
 
 |4 S 
 
 
 h 
 
 
 ^rt S* 
 
 
 o 
 
 * *y 
 
 W3 . sT J; 
 
 ^ 
 
 * s 
 
 ^3 53 
 
 5 -,g c c 
 
 ^ 
 
 
 S " 
 
 & s g 
 
 8 
 
 B 
 
 
 C C Jg n! 
 
 K 
 
 g 
 
 
 3*3 "B^ 
 
 
 
 . i? . . . 
 
 p? 
 
 ^ 
 
 
 
 "3 
 
 
 
 M ^ 
 
 ' | ' ' ' ' | = - - = = 
 
 - - ~ ^ - 
 
 2 
 
 
 
 H H 
 
 H 
 
 H 
 
 
 
 
 8 
 
 
 
 </T t/T 
 
 "* ifi "* * 
 
 T 
 
 | 
 
 u u 
 
 U * U U rH 
 
 a; 
 
 M 
 
 a- - - - tfT o uT 
 .8 a 
 
 cL ~ a a " a 
 
 a 
 
 g 
 
 
 t * * W W W 
 
 in 
 
 S 
 
 ' S" S S* 
 
 * E * S S " 
 
 g 
 
 
 o " = - * S | - - 
 
 o ** o o *" o 
 
 g 
 
 i 
 
 
 fa fa fa fa 
 
 fa 
 
 
 
 
 
 
 g 
 
 Of 
 
 u< 2* 2* 
 
 
 1 
 
 c?. - - - - g ?. - - , 
 
 HJ ^ J 
 
 %-z- - - - - - * 3* 
 
 Jj 
 
 3 
 
 
 
 
 - 
 
 | , | 5 ill 
 
 5 1 i 1 1 1 
 
 4 
 
 9 
 
 > >> >> 
 
 >,>>>, IS fc * 
 
 i 
 
 S 
 
 Q <o g jT <T ( jfS^'""3'v 
 
 nT S uJ" 13 T 13 13 <u <L T 73 -^ 
 
 g 
 
 3 
 
 o o ooooo ' ' 'S 
 
 ' o ' o ' ' a "o '-*- 1 ' o o .? SP 
 
 5 
 
 
 
 S^-<S<<ilS Sc75 w 
 
 
 GO 
 
 ^ 4 - 2* - * 
 
 II c ^ . < ?' C ? * .^l 1 c 
 
 4 * * 
 
 J3 
 
 | 
 
 
 4t>||in<N|l 4| |iO 4 
 
 ^ 
 
 K 
 
 XX XX XX.XXXXXx 
 
 XXXXXXX XX XX X 
 
 x 
 
 ^ 
 
 
 
 U5 
 
 a 
 
 ! rH ffi rH r-i 
 
 <N ! ( - ' OJ G*. *1 
 
 rn 
 
 
 b 1 
 
 J ~ t ^ 7 
 
 ^ 
 
 
 
 oo 
 
 
 
 
 1 
 
 g 
 
 
 09 - 
 
 s ^ 
 
 -s S . - 8 - 
 
 3 
 3 
 
 o 
 
 .2" ? -2 of en .2 'C co" " ^ 
 
 S 5 3 S 3 I .^|| S I 
 
 11 la 1? ill Pi 
 
 03 u^ ed^^BUiJ^u 
 
 <co >S DiSDa)Ct<b33Q 
 
 |3s sls| I? ^-| 
 
 it la ill* ~s 3s8 
 
 2.-lo3g 5^ ^2|S 
 
 DiCUiJN^WOO fc'o JOD 
 
 Fusarium- 
 A.quae-duct 
 
 
 CQ PQ QQPQ rnnrnrornrnrnrn 
 
 
 n 
 
286 
 
 APPENDIX B 
 
 No. 4. LIST OF SPORE-BEARING GERMS FOUND IN WATER. 
 
 GEBM. 
 
 ACTION 
 ON 
 GELATIN. 
 
 ENDS. 
 
 COLOK. 
 
 MORPHOLOGY. 
 
 B. Anthracis, 
 
 Liq., 
 
 Square, 
 
 Gray white, 
 
 Forms filaments, 
 
 
 
 
 
 3-6-10 M x 1-1.5 M. 
 
 B. Subtilis, 
 
 " 
 
 Round, 
 
 Opaque white, 
 
 Long filaments, G ^ x2ju.. 
 
 B. Vermicularis, 
 
 " 
 
 " 
 
 Gray, 
 
 Forms extensive filaments, 
 
 B. Megaterium, 
 
 M 
 
 
 
 Whitish, 
 
 8-9 fji long, 2.5 M broad. 
 
 B. Ramosus, 
 
 " 
 
 " 
 
 Gray, 
 
 Long filaments, 7 /u x 1.7/m. 
 
 B. Mycoides, 
 
 " 
 
 . . 
 
 White, 
 
 Long filaments, 1. 6-2.4 p. x .9 p.. 
 
 B. Tetanus, 
 
 " 
 
 Round, 
 
 . . . 
 
 .9-1.1 M x -1-.2 M. 
 
 B. Pyocyanus, 
 
 " 
 
 
 Greenish white, 
 
 .8-Ux.15-.25M. 
 
 B. Mouse septicaemia, 
 
 " 
 
 
 Yellowish white, 
 
 Frequently in pairs, .8-1 /u. ,l-.2fi. 
 
 B. Brunneus, 
 
 Non-liq., 
 
 . . 
 
 Milk-white brown , 
 
 Fine and slender. 
 
 B. Circulans, 
 
 Liq., 
 
 Round, 
 
 Translucent, 
 
 In twos and fours, 2-5 /u. x 1 M- 
 
 B. Erythrosporus, 
 
 Non-liq., 
 
 " 
 
 Fluor-green, 
 
 Slender and short filaments. 
 
 B. Der Rothe, 
 
 Liq., 
 
 " 
 
 Raspberry red, 
 
 Small, forms filaments. 
 
 B. Cuticularis albus, 
 
 Non-liq., 
 
 " 
 
 White, 
 
 Bent filaments, 3.2 fi long. 
 
 B. Granulosus, 
 
 Liq., 
 
 . . 
 
 Yellowish white, 
 
 Long slender filaments. 
 
 B. Limosus, 
 
 " 
 
 Round, 
 
 White, 
 
 Two or three joined, 3-4 /a xl.25/u.. 
 
 B. Zopfii, 
 
 " 
 
 . . 
 
 Whitish yellow, 
 
 2-5 /ax. 75-1 M. 
 
 B. Mesentericus ruber, 
 
 " 
 
 Round, 
 
 Yellow brown, 
 
 Slender. 
 
 B. Mesentericus fuscus, 
 
 " 
 
 . . 
 
 Brownish yellow, 
 
 Short, in twos and fours. 
 
 B. Mesentericus 
 
 
 
 
 
 vulgatus, 
 
 11 
 
 Round, 
 
 Yellow, 
 
 Small, fat, pairs and fours. 
 
 B. Iridescens, 
 
 " 
 
 . . 
 
 Greenish yellow, 
 
 Bent filaments, 3.5-5.2 /u. long. 
 
 B. Guttatus, 
 
 " 
 
 . . 
 
 Bluish white, 
 
 1-1.13/u. x .93 u. 
 
 B. Thalassophilus, 
 
 " 
 
 . . 
 
 Light gray, 
 
 An anaerobian, with slender, vari- 
 
 
 
 
 
 able filaments. 
 
 B. Amylozyme, 
 
 Non-liq., 
 
 Round, 
 
 White, 
 
 Pairs and chains, 2-3 /j. x -5 /u.. 
 
 B. Filiformis, 
 
 Liq., 
 
 
 White, 
 
 Forms filaments, 4 n. x 1 /<* 
 
 Bacillus "C," 
 
 Liq., 
 
 . . 
 
 Pale brown, 
 
 5-20 nxl/*. 
 
 Bacillus "D," 
 
 Non-liq., 
 
 Round, 
 
 Pearl, 
 
 l-2/utx.l-.2M. 
 
 B. Acidi lactici, 
 
 Non-liq., 
 
 . . 
 
 Gray white, 
 
 Pairs and fours, 1-1.7 M x .3-.4fx. 
 
 B. Lactis cyanogenus, 
 
 " 
 
 Blunted 
 corners, 
 
 Gray, 
 
 l-4 M x.3-.5 M . 
 
 B. Butyricus, 
 
 Liq., 
 
 
 Dirty yellow, 
 
 2.1/HX.38M. 
 
 B. Crassus aromaticus, 
 
 Round, 
 
 White 
 
 3.5-5 /u. x 1.5 /n. 
 
 B. Aerophilus, 
 
 " 
 
 " 
 
 Greenish yellow, 
 
 Slender, in twos and filaments. 
 
 B. Muscoides, 
 
 Non-liq., 
 
 . . 
 
 Opalescent, 
 
 IM- broad. 
 
 B. Putrificus coli, 
 
 Liq., 
 
 . . 
 
 White, 
 
 Slender filaments, 3/u. long. 
 
 B. Thermophilus, 
 
 Non-liq., 
 
 
 White, 
 
 Forms filaments. 
 
APPENDIX C. 
 
 THE LEGAL LIABILITY OF CITIES AND WATER COMPANIES 
 FOR DAMAGES BY SEWAGE POLLUTED WATER. 
 
 EXCEPTING cities are compelled by law to procure water from 
 satisfactory natural sources, or adopt the most perfect methods of 
 water purification, progress in the hygiene of public water supplies 
 will be comparatively slow. When, however, judicial decrees are 
 obtained against vendors (whether municipal corporations or pri- 
 vate companies) for the distribution of polluted and unwholesome 
 waters, then the interest of water purveyors in the quality of their 
 commodities will be great indeed. 
 
 If cities and private companies are held legally responsible 
 for all losses of life, time, and money by reason of polluted public 
 water supplies, the problem then will not be, is improvement 
 in water quality desirable, but rather how can satisfactory im- 
 provement be obtained. The cost will not be seriously debated 
 then ; because the possible loss of money by damage suits, for a 
 brief period of time, will more than balance the cost of water from 
 proper sources, or of the most perfect works for the artificial 
 purification of polluted waters. 
 
 Let^ it be understood that every gallon of water sent through 
 the public mains must carry with it the seal of approval of con- 
 scientious as well as competent water analysts, and our public sup- 
 plies will then come from sources beyond the reach of sewage 
 pollution, or will be brought to the highest state of artificial puri- 
 fication which is attainable. 
 
 A successful suit has recently been fought upon these lines in 
 the lower courts of the State of Wisconsin. There, in the city of 
 Ashland, an epidemic of typhoid fever occurred during the winter 
 
 287 
 
288 APPENDIX C. 
 
 of 1893-94 ; and notwithstanding repeated complaints by the local 
 and state health officials, the Water Company continued to supply 
 a sewage polluted water to its consumers. Among the victims of 
 this epidemic was one Lars G. Green, a laboring man, whose widow, 
 Mrs. Julia L. Green, upon advice of counsel, began a suit against 
 the Ashland Water Company for the legal value of her husband's 
 life. 
 
 The source of water supply for Ashland was Chequamegon 
 Bay, an arm of Lake Superior. The same bay also serves as the 
 receptacle of the city's sewage (Chap. II.) ; and the water supplied 
 to the patrons of the public mains was a mixture of water as re- 
 ceived from natural sources into the bay, and the city's sewage. 
 
 The suit was based on the theory that the water from Che- 
 quamegon Bay contained the specific germs of typhoid fever, which 
 came into it through the city sewers in the dejections of typhoid 
 fever patients then in Ashland ; that this water was drunk by Mr. 
 Green, and laid the foundation for the disease by which he per- 
 ished ; and that the Water Company, knowing the condition of the 
 water in the bay, was negligent in supplying to their customers 
 water for dietetic uses which was sewage polluted and therefore 
 unwholesome. 
 
 The suit was tried in the Circuit Court of Portage County, 
 Wisconsin, during the last week of November (1897), before 
 Hon. Chas. M. Webb. Upon trial of the case, it was proven that 
 Mr. Green was a railway employee living and working continuously 
 in Ashland ; that he was taken ill with and died of typhoid fever ; 
 that his premises were supplied with water from the mains of 
 the Ashland Water Company ; that the only water available in 
 Ashland was that supplied by the water company from Chequame- 
 gon Bay ; that this water was polluted with the sewage from the 
 city of Ashland ; that previous to Mr. Green's illness typhoid fever 
 had prevailed in Ashland, and the dejections from the patients had 
 gone into the city sewers and been discharged into Chequamegon 
 Bay ; and that with the exception of four days just prior to his ill- 
 ness, Mr. Green was exposed to the influence of no other water than 
 that supplied to his premises and the city of Ashland. 
 
 From the testimony offered, the jury found that the typhoid 
 
APPENDIX C. ^W^ K*>^ 289 
 
 germ was transmitted to Mr. Green through the medium of the 
 public water supply, and held the water company liable in $5,000 
 damages. (The legal value of a human life under the laws of 
 Wisconsin.) 
 
 Similar suits doubtless will be brought elsewhere, to settle 
 the question of liability of municipal corporations and water com- 
 panies for delivering to their citizens or customers a fluid which 
 is carrying the germs of dangerous disease. 
 
AUTHORITIES QUOTED OR REFERRED TO. 
 
 ABBOTT, DR. A. C., Philadelphia, " Principles of Bacteriology." 
 
 ALESSI, DR. G., London, " Putrid Gases as Predisposing 
 
 Causes of Typhoid Infection." 
 
 ANDERSON PURIFIER Co., London, 1896, " Water Purification." 
 
 ANKLAM, F., Berlin, " Filters at Lake Miiggel." 
 
 Annales de L? Institute Pasteur, Paris, 1892-94. 
 Annual Summary of Vital Sta- 
 
 London, 1890-96. 
 
 BAKER, M. N., New York, " Mechanical Filters," etc. 
 
 BAUMEISTER, PROF. R., Carlsrahe, "CleaningandSeweragecf Cities." 
 
 BERTSCHINGER, DR. A., Zurich, " Analyses of Zurich Water." 
 
 BINNIE, SIR A. R., London, 1894, " Available Sources of Water Sup- 
 
 ply for London." 
 BLESSING, JAMES H., Albany, N. Y., 1897, " An Address to the Common 
 
 Council." 
 
 BRYAN, W. B., London, " Cleaning Filters." East London 
 
 Water*Company. 
 
 CARMICHAEL, PROF. H., Boston, 1896, " Reduction of Iron in Ground 
 
 Waters." 
 Centralblatt fur Bacteriologie, 1892, " Freudenrich's Tests of Pasteur 
 
 Filters." 
 Consular Reports, U.S., Washington, D.C., " Fischer Filter, Worms." 
 
 1897, 
 CROOKSHANK, E. M., London, " Manual of Bacteriology." 
 
 DE VARONA, I. M., Brooklyn, 1896, " Report on the Future Extension 
 
 of the Water Supply of Brook- 
 lyn." 
 
 DEVONSHIRE, E., London, "Anderson Revolving Iron Puri- 
 
 fier." 
 
 DIBDIN, W. J. London, " Analytical Investigation of Lon- 
 
 don Waters." 
 
 DROWN, DR. T. M., Lehigh University, " Reduction of Iron in Ground 
 
 South Bethlehem, Waters," etc. 
 Pa., 
 DUNBAR, PROFESSOR DR., Hamburg, " Reduction of Iron in Ground 
 
 Waters," etc. 
 290 
 
AUTHORITIES QUOTED OR REFERRED TO. 
 
 291 
 
 DUPRE, DR. A., 
 
 London, 
 
 ENGINEER Commission on Extension and Better- 
 ment of Cincinnati Water Works, 1896 : 
 Engineering News, 
 
 Engineering Record, 
 ERNST, DR. H. C., 
 
 Fire and Water, 
 FLAD, EDWARD, 
 FRANKLAND, DR. E., 
 
 FRANKLAND, PROF. PERCY 
 and G. C., 
 
 GILL, HENRY,* 
 
 " Andersfcn Iron Purifier, Worces 
 ter, England." 
 
 " Report" on Filtration of Water." 
 New York, 1896, " Removal of Iron from Ground 
 
 Waters," etc. 
 
 New York, 1894, " Typhoid Fever Statistics." 
 Harvard University, " Examination of Sample of Well 
 Boston, Wa.ter." 
 
 New York, 1804-95, " Typhoid Fever Statistics." 
 St. Louis, "Subsidence of Ohio River Water." 
 
 London, " Operation of London Filters," 
 
 etc. 
 
 Birmingham, Eng., " Micro Organisms in Water." 
 
 HAWKSLEY, THOS., 
 HAZEN, ALLEN, 
 
 HOLLIS, F. S., 
 
 JORDAN, DR. EDWIN O., 
 
 London, 
 
 London, 
 New York, 
 
 Boston, 
 
 Chicago, 
 Karlsruhe, 1897, 
 
 Journal fur Gasbeleuchtung 
 
 und Wasserversorgung, 
 
 Journal of the Sanitary Institiite, London, 1894-95. 
 Journal of the N. E. W. W. As- 
 
 1896. 
 
 New York, 
 London, 
 Altona, Ger., 
 
 London, 
 
 KIRKWOOD, J. P.,* 
 KLEIN, DR. E., 
 
 KUMMEL, W.,* 
 
 LANDOIS-STERLING, 
 LANDRETH, PROF. O. H., 
 
 "Filtration of the Muggel Lake 
 Water Supply." 
 
 " Area of London Filters." 
 " Filtration of Public Water Sup- 
 plies," etc. 
 
 " Bacterial Efficiency of Lorain 
 Filters." 
 
 " Identification of Typhoid Fever 
 
 Bacillus." 
 " Cleaning Sand Filters under 
 
 Ice." 
 
 " Filtration," etc. 
 " Influence of Light on Micro 
 Organisms." 
 
 " Filtration of River Waters." 
 " Quality of London Waters." 
 " Rates of Filtration." 
 
 " Human Physiology." 
 
 Schenectady, N.Y., " Epidemic of Typhoid Fever at 
 Elmira, N.Y." 
 
 LANKESTER, PROF. E. RAY, Oxford University, "Origin of b. typhosus" etc. 
 
 LEFFMANN, DR. HENRY, 
 MAGER, ED., 
 
 Philadelphia, 1897, 
 
 Hamburg, 1897, 
 
 Unfiltered Surface Waters al- 
 ways Unsafe for Town Supply." 
 
 Process of Cleaning the open 
 Filters of the Hamburg Water 
 Works." 
 
 Deceased. 
 
292 
 
 AUTHORITIES QUOTED OR REFERRED TO. 
 
 Manual of American Water Works, New York, 1897- 
 
 MASON, PROF. WILLIAM P., New York, 1896, 
 
 MASS. State Board of Health, Boston, 1890-95, 
 
 MCFARLAND, DR. Jos., Philadelphia, 
 
 MEYER, F. ANDREAS, Hamburg, 1894, 
 
 MIGULA, Dr. W M 
 MILLS, H. F., 
 MIQUEL, DR. PIERRE, 
 MOORE, DR. V. A. 
 
 London, 1893, 
 Lawrence, Mass., 
 Paris, 
 Washington, D.C., 
 
 MORISON-jEWELL FILTRATION COMPANY, 
 
 MUNN, DR. WILLIAM P., 
 
 NICHOLS, WM. RIPLEY, 
 
 ODLING, DR. WILLIAM, 
 Ohio Sanitary Bulletin, 
 OSLER, DR. WILLIAM, 
 
 Denver, Col., 1896, 
 
 " Water Supply." 
 
 Annual Reports. 
 
 " The Pathogenic Bacteria." 
 
 " Das Wasserwerk der Frien- und 
 
 Hansestadt, Hamburg." 
 " Practical Bacteriology." 
 " Sterilization of Filter Sand." 
 " Sedimentation of Waters," etc. 
 " Character of the Flagella." 
 " Experimental Filters for Phila- 
 delphia," 1897, etc. 
 " Preliminary Report of Health 
 Commissioner." 
 
 New York, 1883, " Water Supply.' 
 
 London, 1893, 
 Ohio, 1897, 
 Baltimore, 
 
 PETTENKOFER, DR. MAX VON, Munich, 
 PIEFKE, HERR, Berlin, 
 
 PRELLER, CHARLES, S. D., London, 
 
 " Filtration of Surface Waters." 
 
 " Lorain Filters." 
 
 " Typhoid Fever in Baltimore." 
 
 " Cause of Typhoid Fever." 
 
 " Die Principien der Reinwasserge- 
 winnung vermittelst Filtration." 
 
 "Water Works of Zurich, Swit- 
 zerland." 
 
 "Use of Sterilized Water at 
 World's Fair." (1893.) 
 
 Proceedings American Water- 
 Works Association, 1894, 
 
 Proceedings Institution of Civil 
 
 Engineers, London, 1892-1895, " Filtration of Water in Europe." 
 
 PRUDDEN, DR. T. M., New York, " Typhoid Bacillus in Water." 
 
 RAFTER AND MALLORY, 
 RAVENEL, DR. M. P., 
 REINCKE, DR. J. J., 
 
 ROGERS, DR. EDMUND, 
 ROSENAU, Dr. M. J , 
 ROYAL COMMISSION, 
 
 SANARELLI, Dr. G., 
 
 SCHRODER, RUD, 
 SEDGWICK, PROF. W. T., 
 
 SHEDD, J. HERBERT* 
 
 Rochester, N.Y., " Report on Spring Water Epi- 
 demic." 
 Philadelphia, 1897, " Bacterial Tests of Chemung 
 
 Water, Elmira, N.Y. 
 Hamburg, " Epidemiology of Typhoid Fever 
 
 in Hamburg and Altona." 
 
 Denver, Col., " Cause of Mountain Fever. " 
 
 San Francisco, " San Francisco Water Supply." 
 
 London, 1893, "Metropolitan Water Supply." 
 
 Montevideo, S.A., "The Typhoid Bacillus and Eti- 
 ology of Typhoid Fever." 
 
 Hamburg, " Operation of Hamburg Filters." 
 
 Boston, " Bacteria in Spring and Well 
 
 Waters," etc. 
 
 Providence, R.I., " Sand Filtration." 
 
AUTHORITIES QUOTED OR REFERRED TO. 
 
 293 
 
 SMITH, DR. THEOBALD, Boston, 
 
 Statistische Zusam menstellung 
 
 der Betrtebs Ergebnisse von 
 
 Wasserwerken, Munich, 1895. 
 
 STERNBERG, DR. GEO. M., Washington, D.C. 
 
 THOMAS, R. J., Lowell, Mass., 
 
 THOMAS AND MARSHALL, DRS., Philadelphia, 
 
 THORNE, DR. THORNE, 
 
 London, 
 
 New York, 1894, 
 
 Times, Daily, 
 
 Transactions American Society 
 
 of Civil Engineers, New York, 
 
 Tribune, Daily, New York, 1894, 
 
 VAUGHAN, DR. V. C., 
 VAUGHAN AND Now, 
 
 Ann Arbor, Mich., 
 Philadelphia, 
 
 The Fermentation Tube.' 
 
 " Sources of German Water Sup 
 
 plies." 
 " Manual of Bacteriology." 
 
 " Water Supply of Lowell." 
 
 Report on Philadelphia Water 
 Supply. 
 
 " The Caterham and Redhill Epi- 
 demic." 
 
 " Typhoid Fever." 
 
 Vols. xxxi., xxxii., xxxiii., and xxxv. 
 " Typhoid Fever and Water Sup- 
 ply." 
 
 " A Bacteriological Study of 
 
 Drinking Water." 
 " The Ptomaines and Leuco- 
 
 WESTON, EDMUND B., 
 WHIFFLE, G. C., 
 
 Providence, R.I., 
 Boston, 
 
 WOODHEAD, DR. G. SIMS, Edinburgh, 
 
 ZEIMSSEN, PROFESSOR VON, Munich, 
 ' itschriftdes Vereines Deutsch- 
 
 & Ingenieurie, Hamburg, 1895, 
 
 <_ 'thrift fur Hygiene, 1896, 
 
 " Providence Experimental Filter 
 Tests." 
 
 " Some Observations on the Re- 
 lation of Light to the Growth 
 of Diatoms." 
 
 " Seasonal Distribution of Ty- 
 phoid Fever," etc. 
 
 " On Typhoid Fever in Munich." 
 " Apparatus for Washing Filter 
 
 Sand." 
 " Reduction of Iron in Ground 
 
 Waters." 
 
INDEX. 
 
 PAGE 
 
 Action of b. typhosus on animals ... 63 
 Action of bacteria on organic matter .... 132 
 Action in water of metallic iron on suspended 
 
 matter 241 
 
 Advantage of sedimentation 246 
 
 Advantage of water from undefiled sources . 83 
 
 Aeration of filter-bed 7, 133, 137, 235 
 
 Aeration of polluted waters 21 
 
 Aeration of water by Pohle air lift .... 204 
 Aeration of water from Anderson Purifier . . 241 
 
 Aerobic bacteria 274 
 
 Air-borne diseases 37 
 
 Altona, open filters of 252 
 
 Alum and sand filtration 173 
 
 Alum consumption per gallon of water with 
 
 mechanical filters 201 
 
 Alum per gallon of water, Lorain, Ohio . . . 196 
 Alum per gallon of water, Providence, R.I. . 189 
 
 Alum, reduction of bacteria by 115 
 
 Alum, test for, in filtrate 189 
 
 Alum, uudecompossd, in subsided water . . 115 
 Alum, use of, in Philadelphia water supply . 201 
 Alum, variable consumption of, with mechani- 
 cal filters 201 
 
 Amsterdam, filters of 181 
 
 Anaerobic bacteria 274 
 
 Analytical investigation of London water . .117 
 Analysis of Reading, Mass., water . . . .207 
 
 Analysis of sands and gravels 162 
 
 Analysis of solids of putrefactive bacteria . . 272 
 
 Analysis of sulphate of alumina 189 
 
 Analysis of water at Long Branch, N.J. . . 195 
 Anderson Revolving Iron Purifier, 
 
 52, 156, 182, 241, 242, 243, 244, 245 
 
 Angers, France, filter gallery 179 
 
 Animal charcoal filters for reduction of iron, 
 
 205,206 
 Anklam and Oesten's method for reduction of 
 
 iron in ground waters 205 
 
 Anthrax protein 273 
 
 Apparatus required for sterilization of water . 123 
 
 Area of filters, Amsterdam 181 
 
 Area of filters, The Hague 181 
 
 Area of filters, Hamburg 210, 223 
 
 Area of filters, Hudson, N.Y 177 
 
 Area of niters, Lawrence, Mass. . . . 166, 169 
 
 PAGE 
 
 Area of filters, London 183 
 
 Area of filters, Lowell, Mass 176 
 
 Area of filters, Miiggel Lake 231 
 
 Area of filters, Paris 182 
 
 Area of filters, Poughkeepsie, N.Y 178 
 
 Area of filters, Rotterdam 180 
 
 Area of filters, Zurich 183 
 
 Area of water on sand-bed, Cincinnati filters . 249 
 Arrangement of filtering materials, Amsterdam, 181 
 Arrangement of filtering materials, Berlin . . 230 
 Arrangement of filtering materials, Cincinnati, 250 
 Arrangement of filtering materials, The Hague, 181 
 Arrangement of filtering materials, Hamburg, 213 
 Arrangement of filtering materials, Hudson, 
 
 N.Y 177 
 
 Arrangement of filtering materials, Lawrence, 
 
 Mass 166 
 
 Arrangement of filtering materials, Lowell, 
 
 Mass 175 
 
 Arrangement of filtering materials, Miiggel 
 
 Lake 232 
 
 Arrangement of filtering materials, Paris . . 182 
 Arrangement of filtering materials, Pough- 
 keepsie, N.Y 178 
 
 Arrangement of filtering materials, Rotterdam, 180 
 Arrangement of filtering materials, Zurich . . 183 
 Arrangement of main and lateral drains . 249, 250 
 
 Artesian well waters 102, 113 
 
 Arthrospores 280 
 
 Asbury Park, N.J., mechanical filters, 203, 263, 264 
 Ashland, Wis., sand filters . . . 261,262,265 
 Assimilation of minerals in water by plant life, 88 
 
 Asterionella 87 
 
 Average daily consumption of water, Amster- 
 dam 182 
 
 Average daily consumption of water, Berlin . 233 
 Average daily consumption of water, The 
 
 Hague 181 
 
 Average daily consumption of water, Ham- 
 burg 220 
 
 Average daily consumption of water, Law- 
 rence, Mass 169 
 
 Average daily consumption of water, Munich, 80 
 Average daily consumption of water, Rotter- 
 dam 180 
 
 Average daily consumption of water, Zurich . 183 
 
 295 
 
296 
 
 INDEX. 
 
 PAGE 
 
 Average length of run of mechanical filters . 191 
 Average period of operation of filters, Ham- 
 burg 220 
 
 Average rate, Amsterdam filters 182 
 
 Average rate, Berlin filters 260 
 
 Average rate, The Hague niters 181 
 
 Average rate, Hamburg filters 260 
 
 Average rate, Lawrence, Mass., filters . 165, 169 
 
 Average rate, London filters 259 
 
 Average rate, Paris suburbs filters .... 182 
 
 Average rate, Rotterdam filters 180 
 
 Average rate, Somerville and Raritan, N.J., fil- 
 ters 194 
 
 Average rate, Zurich filters 260 
 
 Average size of sand-grains 164 
 
 Bacilli, The 275, 277 
 
 Bacteria, The 272 
 
 Bacteria found in water larger than b. typhosus, 285 
 Bacteria found in water resembling b. typhosits, 283 
 Bacteria found in water smaller than b. typhosus, 284 
 
 Bacteria and organic matter 273 
 
 Bacteria from human intestine 60 
 
 Bacteria in air and soil 279 
 
 Bacteria in air of laboratory 53 
 
 Bacteria in artificial ice 53 
 
 Bacteria in Chelsea water 144 
 
 Bacteria in cistern water 41 
 
 Bacteria in distilled water 52 
 
 Bacteria in East London water 147 
 
 Bacteria in Elbe water 144, 152 
 
 Bacteria in filtered Elbe water 152 
 
 Bacteria in Grand Junction water 146 
 
 Bacteria in Lambeth water 146 
 
 Bacteria in New River water 147 
 
 Bacteria in Ohio River water 41 
 
 Bacteria in rain-water 50, 51, 279 
 
 Bacteria in spring water 48 
 
 Bacteria in Southwark and Vauxhall water . 145 
 Bacteria in water from stone disk and tube 
 
 filters 47 
 
 Bacteria in well waters ... 22, 26, 27, 44, 45 
 Bacteria in West Middlesex water .... 145 
 Bacteria, pathogenic ... 2, 15, 103, 278, 280 
 
 Bacteria, putrefactive 2, 278, 279 
 
 Bacteria, reduction of, by alum . . .. . 112,115 
 Bacteria, reduction of, by lime . . . 112, 114, 117 
 Bacteria, reduction of, by sterilized clay . .114 
 Bacteria, reduction of, by subsidence . Ill, 148 
 Bacteria, reduction of, by subsidence and filtra- 
 tion 148 
 
 Bacterial action of metallic iron in water, 241, 243 
 Bacterial contents of Chemung River water, 
 
 100, 101, 198, 199 
 
 Bacterial contents of Vanne water 243 
 
 Bacterial Contents of Various Waters . 40 
 Bacterial efficiency, Anderson Purifier, Paris . 243 
 Bacterial efficiency of filters at high rates, 
 
 173, 175, 190, 199, 202 
 
 PAGE 
 
 Bacterial efficiency of filters, Berlin .... 237 
 Bacterial efficiency of filters, Hamburg . 217, 218 
 Bacterial efficiency of filters, Lawrence, Mass., 
 
 143, 153 
 Bacterial efficiency of filters, Lorain, Ohio, 
 
 195, 196, 202 
 
 Bacterial efficiency of filters, Paris .... 182 
 Bacterial efficiency of filters, Rotterdam . . 180 
 Bacterial efficiency of Morison mechanical fil- 
 ter 172, 199 
 
 Bacterial efficiency of sand filters 143 
 
 B. aiithracis 273, 276, 280, 281, 285 
 
 B. colt comntunis 275, 277, 283 
 
 B, coli comtmtnis, and b. lactis aerogenes, pro- 
 duction of gas by 57, 58, 
 
 B. coli communis and typhoid fever .... 63 
 B. coli communis in Chemung River water . 101 
 
 B. coli comntunis in dejecta 15 
 
 B. coli communis in impounded water . . . 103 
 B. coli communis, inoculation of guinea pigs 
 
 with 61 
 
 B. coli communis in polluted waters .... 60 
 B. coli communis in water supply of San Fran- 
 cisco 28 
 
 B. coli communis in well water .... 28, 103 
 
 B. diphtheria 275, 276, 278, 281 
 
 B. diphtheria, no proof of, in water .... 281 
 
 B. fluorescent liquefaciens 277, 283 
 
 B, lactis aerogenes . 57, 58, 60, 277, 280, 282, 283 
 
 B. mesentericus -vulgatus 279, 286 
 
 B, of mouse septiccemia 280, 284 
 
 B. prodigiosus 277, 285 
 
 B. proteus fluorescens 278, 280, 282 
 
 B. proteus vulgaris .... 28, 103, 279, 280 
 B. protetis vulgaris in water supply of San 
 
 Francisco 28 
 
 B. pyocyanus 280, 284 
 
 B. rothe 277, 286 
 
 B. tetanus 274,280,281,284,286 
 
 B. tuberculosis 273, 275, 280, 281 
 
 B. typhosns . 9, 11, 13, 15, 17, 30, 49, 56, 57, 58, 59, 
 60, 61, 62, G3, 64, 65, 275, 276, 277, 278, 280, 
 281, 282, 283 
 B. typhosus, b. coli communis, and b. lactis 
 
 aerogenes 57,58,59,60 
 
 B. typhosus, chemical composition of . . . 56, 57 
 B. typhosus, form and dimensions of . . 57, 278 
 
 B. typhosus in acid solution 58 
 
 B. typhosus in fermentation tube 57 
 
 B. typhosus, influence of, on gelatin, 56, 57, 58, 283 
 B. typhosus in sterilized milk . . . . 56, 58, 59 
 
 B. venenosus 99 
 
 B. violaceus 278 
 
 Beer, consumption of, in European cities, 
 
 13, 78, 79, 80, 139 
 
 Berlin and Chicago, water consumption com- 
 pared 8 
 
 Berlin, Plain Sand Filters of 230 
 
 Berlin, sources of water supply 75 
 
INDEX. 
 
 297 
 
 PAGE 
 
 Biologic action of sand filters . 132, 133, 154, 165 
 Boiled water ............ 121 
 
 Brieger, reduction of toxins in bacterial cul- 
 tures ............ 3, 62 
 
 Brooklyn, ground water supply ..... 82 
 
 Capacity of Anderson Purifiers, Paris . . 243 
 Capacity of clear-well, Cincinnati niters . 250, 251 
 Capacity of filters, Mtiggel Lake . . . 231, 234 
 Capacity of Fischer filter, Worms ..... 240 
 
 Capacity of Lawrence, Mass., City filter . . 168 
 Capacity of London filters ....... 183 
 
 Capacity of mechanical filters, Somerville and 
 Raritan, N.J ........... 194 
 
 Capacity of proposed filters, Cincinnati . . . 247 
 Capacity of proposed settling-basins, Cincin- 
 nati .............. 247 
 
 Capacity of Schroder sand washer ..... 223 
 
 . 208 
 . 93 
 . 25 
 2, 85 
 135 
 56 
 
 Capacity of settling-basins, Hamburg . . 
 Caterham and Redhill, epidemic of typhoid 
 Causes of failure in filtration of water .. 
 Changes in the quality of a water supply . 
 Changes 'of pressure on sand-bed 
 Characteristics of b, typhosus 
 Chelsea filters, London ...... 25, 144 
 
 Chemical analysis of Lake Superior water . . 99 
 Chemical and bacterial efficiency of Fischer 
 
 filter ............. 240 
 
 Chemical and biological changes by sedimenta- 
 
 tion .............. 110 
 
 Chemical composition of b. typhostts .... 56 
 
 Chemical composition of the bacteria . . . 272 
 Chemical constituents of rain-water .... 51 
 
 Chemically and bacterially pure water ... 3 
 Chemicals used in Morison experimental filter, 
 
 187, 188 
 Chemung River as a carrier of typhoid infec- 
 
 tion .............. 100 
 
 Chicago and Berlin, water consumption com- 
 
 pared ............. 82 
 
 Chlorophyl ............. 272 
 
 Cholera bacillus ........ 23,276,280 
 
 Chromogenic bacteria ......... 277 
 
 Cincinnati and Hamburg, water consumption 
 
 compared ............ 82 
 
 Cincinnati as a typhoid fever center .... 42 
 
 Cincinnati, filters proposed for ...... 246 
 
 Cisterns, precautions with underground ... 86 
 Cistern water for dietetic purposes .... 86 
 
 Citations on Typhoid Fever Epidemics . 91 
 Cities of second class in United States ... 75 
 Clarified and purified water ....... 46 
 
 Classification of Cities by Typhoid 
 
 Fever Rates .... 70, 71, 72, 73, 74, 75 
 
 Clay sterilized, reduction of bacteria by ... 114 
 Cleaning the sand of mechanical filters . . . 184 
 Clear-water conduit and basin, Hamburg . . 219 
 Clear-water reservoir, Berlin ...... 233 
 
 Coagulants, use of, with mechanical filters . . 184 
 
 PAGE 
 
 Cocci, The 275 
 
 Collection of water in impounding reservoirs . 14 
 Colne Valley water, treatment of with lime . 113 
 
 Comma Bacillus, Koch's 276, 280 
 
 Comparison of cities upon typhoid fever rates, 
 
 Europe and United States ...... 75 
 
 Comparison of continuous and intermittent 
 
 sand filters 143 
 
 Comparison of large and small cities for water 
 
 supply 88 
 
 Comparison of natural and artificial sand fil- 
 tration 138 
 
 Comparison of population and water consump- 
 tion 87 
 
 Comparison of typhoid fever rates, Berlin, 
 Chicago, Louisville, Munich, Pittsburg, 
 
 Vienna 12 
 
 Comparison of water supplies, Europe and 
 
 United States 77 
 
 Constituents of sewage 10 
 
 Construction of filters, Hamburg 210 
 
 Construction of mechanical filters 185 
 
 Construction of plain sand filters 160 
 
 Consumption and waste of water in American 
 
 cities 82 
 
 Consumption of alum with mechanical filters . 201 
 Consumption of beer in American and foreign 
 
 cities 13 
 
 Consumption of beer in Munich 80 
 
 Consumption of water, Amsterdam .... 182 
 
 Consumption of water, Berlin 234 
 
 Consumption of water, German cities ... 84 
 Consumption of water, The Hague .... 181 
 
 Consumption of water, Hamburg 220 
 
 Consumption of water, Jersey City, N.J. . . 34 
 Consumption of water, Lawrence, Mass. . . 169 
 
 Consumption of water, London 118 
 
 Consumption of water, Munich 80 
 
 Consumption of water, Rotterdam 180 
 
 Consumption of water, Zurich 183 
 
 Continuous sand filtration 158 
 
 Continuous use of sand filters, London . . . 261 
 Cost of Anderson Revolving Iron Purifier, 243, 244 
 
 Cost of Ashland, Wis., filters 259 
 
 Cost of cleaning Lowell, Mass., filter . . . 176 
 
 Cost of covered filters, Berlin 258 
 
 Cost of covered filters, Zurich 258 
 
 Cost of Filters and Filtration 255 
 
 Cost of filters, Berlin and Hamburg . . 253, 254 
 
 Cost of filters, Long Branch 195 
 
 Cost of filters, Somerville and Raritan, N.J. . 194 
 
 Cost of filtration for Albany, N.Y 196 
 
 Cost of filtration, Miiggel Lake 233 
 
 Cost of filtration, Poughkeepsie, N.Y. . . .178 
 
 Cost of Fischer filter, Worms 240 
 
 Cost of Kronke method for reduction of iron 
 
 in ground waters 205 
 
 Cost of lime treatment per million U. S. gal- 
 lons . . 119 
 
298 
 
 INDEX. 
 
 Cost of mechanical and plain sand niters, 192, 193 
 Cost of operating Lawrence, Mass., City filter, 168 
 Cost of operating mechanical filters . . 193, 197 
 Cost of operating open and closed filters, Zurich, 
 
 233,264 
 
 Cost of operating plain sand filters . . 197, 263 
 Cost of plain sand filters, 
 
 240, 245, 255, 256, 257, 258 
 
 Cost of Schroder sand washer 223 
 
 Cost of scraping sand-bed, London filters, 
 
 176, 262, 263 
 Cost of sterilizing large quantities of water, 
 
 121, 124, 125 
 Cost of treating water by Anderson process, 
 
 244,245 
 
 Cost of uncovered filters, Berlin and Ham- 
 burg 258 
 
 Cost of uncovered filters, Lawrence, Mass. . . 258 
 
 Cost of uncovered filters, Zurich 258 
 
 Cost of washing sand 263 
 
 Cost of washing sand, Hamburg 264 
 
 Cost of washing sand, Lawrence, Mass. . . 263 
 Cost of water sterilizing apparatus . . 124, 125 
 
 Covered filters 158,232 
 
 Covered reservoirs forstorage of ground waters, 87 
 Covered reservoirs for filtered water . . 156, 157 
 Crenothrix 87 
 
 Dayton, Ohio, Ground Water Supply ... 82 
 Decline of typhoid fever in Vienna and Munich, 67 
 Decomposition of organic matter in water . . 7 
 Degree of purity of natural waters .... 81 
 
 Dejecta from typhoid patients 63, 91 
 
 Delivery of water on sand-bed 250 
 
 Denver, Col., typhoid fever rates . . . 105, 106 
 
 Deposit on sand-bed. (Berlin.) 235 
 
 Digestion, effect of astringent on 200 
 
 Dilution of sewage-polluted water 3 
 
 Dimensions of filters, Cincinnati 247 
 
 Dimensions of filters, Lawrence, Mass. . . . 166 
 
 Dimensions of typhoid bacillus 278 
 
 Diplococci, The 275 
 
 Diphtheria, no proof of infection by water . . 281 
 
 Diseases from soil and water 37 
 
 Distillation of public water supplies . . . 121 
 Distilled water, influence of, on health ... 85 
 
 Distilled water, organic matter in 17 
 
 Distributing reservoirs for filtered water . . 156 
 Distrust of natural sources of water supply . 29 
 
 Domestic filters, use of 120, 131 
 
 Domestic filters, use of, with Ohio River water, 
 
 42, 43, 47 
 
 Domestic use of polluted waters 4 
 
 Dortmund, Ground Water Supply .... 83 
 
 Double filtration of water 154 
 
 Dual system of water supply .... 83, 129 
 Duplicate system of street mains . 121, 124, 125 
 
 Eberth's bacillus 62, 64 
 
 PAGE 
 
 Effective sedimentation of polluted waters . .110 
 Effective size of sand grains .... 143, 164, 172 
 
 Effect of astringent on digestion 200 
 
 Efficiency of Anderson Purifier on Ohio River 
 
 water 52, 244 
 
 Efficiency of filters at high rates of percolation, 2CO 
 
 Efficiency of filters, Hamburg 153 
 
 Efficiency of filters, Lawrence, Mass. . . . 153 
 
 Efficiency of filters, London 148, 151 
 
 Efficiency of filters, measurement of .... 155 
 
 Efficiency of filters, Reading, Mass 207 
 
 Efficiency of filters, Rotterdam 180 
 
 Efficiency of filters, Somerville and Raritan, 
 
 NJ 194 
 
 Efficiency of Kronke method for iron reduction, 205 
 
 Effluent regulating weir 214 
 
 Elizabeth, N.J., epidemic of typhoid . . . 108 
 Elmira, N.Y., epidemic of typhoid .... 100 
 
 Elmira, N.Y., Filtration Works 198 
 
 English Engineers, views of, on sedimenta- 
 tion 141 
 
 Environment, influence of, on species ... 64 
 Epidemic of typhoid, Caterham and Redhill . 93 
 Epidemic of typhoid, Elizabeth, N.J. . . .108 
 
 Epidemic of typhoid, Elmira, N.Y 100 
 
 Epidemic of typhoid, Evansville, Ind. . . . 108 
 
 Epidemic of typhoid, Lausen 91 
 
 Epidemic of typhoid, Lawrence, Mass. ... 98 
 Epidemic of typhoid, Lowell, Mass. ... 98 
 Epidemic of typhoid, Middletown, Conn. . . 106 
 Epidemic of typhoid, Plymouth, Pa. ... 94 
 Epidemic of typhoid, Sault Ste. Marie, Mich. 99 
 Epidemic of typhoid, Spring Water, N.Y. . . 96 
 Epidemic of typhoid, Stamford, Conn. . . . 107 
 Epidemic of typhoid, St. Louis, Mo. ... 99 
 
 Epidemic of typhoid, Zurich 96 
 
 Estimated cost of a system of filters .... 256 
 
 Estimated cost of covered filters 256 
 
 Estimated cost of filters, Albany, N.Y. . . . 257 
 Estimated cost of filters, Cincinnati . . 252, 253 
 Estimated cost of filters, Philadelphia, Pa. . .255 
 Estimated cost of filters, Providence, R.I. . .256 
 
 Estimated cost of open filters 256 
 
 Evansville, Ind., epidemic of typhoid . . . 108 
 Exclusion of b. typhosus from drinking-water . 65 
 Experimental filters, Providence, R.I. . . . 170 
 Experiments on b, typhosus and b. colicommunis 
 
 in sterilized milk 59 
 
 Experiments with sterilized sand, Berlin, 236, 237 
 Evidence of b- typhosus in water supply . . 57 
 Evidence that typhoid fever rates can be con- 
 trolled 75 
 
 Factors of sand filtration 8 
 
 Facultative parasites 273 
 
 Facultative saprophytes 273 
 
 Feasibility of water supply from sources nat- 
 urally pure 90 
 
 Fermentation tube, the 57, 103 
 
INDEX. 
 
 299 
 
 PAGE 
 
 Ferrous hydrate, formation of, by Anderson 
 
 process 241, 243 
 
 Filaments of bacteria 276 
 
 Filter capacity in reserve 210, 231 
 
 Filter galleries 179 
 
 Filter gallery, Angers, France 179 
 
 Filter galleries, Lyons, France 179 
 
 Filter gallery, Perth, Scotland 179 
 
 Filter galleries, rate of percolation .... 179 
 
 Filter gallery, Toulouse, France 179 
 
 Filter galleries, quality of water from . . . 179 
 
 Filters of Hamburg 77, 208 
 
 Filters of the cities of Holland 5 
 
 Filters, management of 83 
 
 Filters, management of, at Berlin 234 
 
 Filters Proposed for Cincinnati . . . .246 
 
 Filters of Zurich 183 
 
 Filters, rate of delivery 161 
 
 Filtered water, use of, for washing sand . . 214 
 
 Filtering materials 162 
 
 Filtration of surface waters 84 
 
 Filtration of lake waters .../..... 20 
 
 Filtration of the Vyrnwy water 81 
 
 Filtration of Water Supplies 131 
 
 Filtration of water from Welsh sources for 
 
 London 84, 141 
 
 Filtration through the drift . . 26, 27, 28, 139, 179 
 
 Filtration with and without alum 173 
 
 Filtration Works, Elmira, N.Y 198 
 
 Filtration Works, Miiggel Lake . . . 230, 231 
 
 Fischer Filter, Worms 238 
 
 Flagella, The, 277 
 
 Food for bacteria in water 134 
 
 Form and dimensions of b. typhosus . . 57, 278 
 Formation of nitrates and nitrites, 133, 136, 137, 194 
 Free sulphuric acid in water treated with 
 
 alum 200 
 
 Freudenrich's tests of Pasteur niters . . . 43, 45 
 
 Gas, production of, by b. colt communis and b. 
 
 lactis aerogenes 58 
 
 Gelatinous film on sand-bed 133 
 
 General use of unpotable water 120 
 
 Germ theory of disease 10 
 
 Grada of sand in London filters ..;... 184 
 
 Grading of filtering materials 162 
 
 Grouping of species 275, 276 
 
 Ground water sources in Germany . . . 83, 84 
 Ground water supply, Brooklyn, N.Y. ... 82 
 Ground water supply, cities of Germany . . 84 
 Ground water supply, Dayton, Ohio .... 82 
 
 Ground water supply, Dortmund 83 
 
 Ground water supply, Jacksonville, Fla. . . 82 
 Ground water supply, Kent Works, London . 83 
 
 Ground water supply, Leipsic 83 
 
 Ground water supply, Lowell, Mass. ... 82 
 Ground water supply, Memphis, Tenn. ... 82 
 Ground water supply, South Bend, Ind. . . 82 
 Growth of bacilli by fission 275, 276 
 
 Growth of bacteria in sand-bed 
 
 PAGE 
 . 134 
 
 Habitat of b. typhosus 64 
 
 Hague, The, filters of 181 
 
 Hamburg and Cincinnati, water consumption 
 
 compared . 82 
 
 Hamburg Settling-Basins and Filters . 208 
 
 Hamburg, water supply of 77, 208 
 
 Hard and soft pure water 31 
 
 Hardness of Ohio River water 28 
 
 Head of water on Berlin filters 232 
 
 Head of water on Fischer filter 238 
 
 Head of water on Hamburg filters .... 213 
 Head of water used on sand washers . . . 223 
 
 High rates of filtration 25, 26, 259, 260 
 
 Holland, filters of 5 
 
 Hollow porous glass plaques for filtration . . 238 
 
 Hudson, N.Y., sand filters 177 
 
 Hygienic Laboratory, Hamburg 5 
 
 Ice, formation of, on filters 158 
 
 Identification of the typhoid bacillus . . 64, 288 
 Imperial Board of Health (Germany) on rate 
 
 of filtration 142 
 
 Impounding reservoirs at high elevations . . 15 
 Impounding reservoirs of New York and Liver- 
 pool 21 
 
 Infection of milk, by typhoid tainted water, 
 
 107, 108 
 
 Infectious disease, sources of 37 
 
 Influence of alkalinity on bacteria in water . 23 
 Influence of b. typhosus on gelatin . . .56, 58 
 Influence of casing-pipe on bacteria in well 
 
 waters 27 
 
 Influence of climate on filter construction . . 251 
 Influence of culture media on species . . . 277 
 Influence of days of cultivation on bacteria . 48 
 Influence of distilled water on bone formation, 86 
 Influence of distilled water on health ... 85 
 Influence of environment on species . . 64, 277 
 Influence of filtered water on bacteria in water 
 
 mains 153 
 
 Influence on filtration of effective size of sand 
 
 grains 173 
 
 Influence of form of filter 158 
 
 Influence of freezing weather on open filters . 149 
 Influence of hard water on the animal system . 30 
 Influence of London water on typhoid fever 
 
 rates 11 
 
 Influence of management on filtration, 83, 253, 254 
 Influence of nitrates and nitrites on bacteria . 23 
 
 Influence of origin on culture 277 
 
 Influence of putrid gases on rats, guinea pigs, 
 
 and rabbits 61 
 
 Influence of sand filtration on typhoid fever 
 
 rates 131, 155 
 
 Influence of sewerage on typhoid rates ... 68 
 Influence of sterilized water on typhoid rates . 128 
 Influence of sunlight on growth of algae . . 87 
 
300 
 
 INDEX. 
 
 PAGE 
 
 Influence of sunlight on growth of bacteria in 
 
 water 53,54,55 
 
 Influence of sunlight on turbid waters ... 87 
 Influence of temperature on growth of species, 277 
 Influence of temperature on growth of bacteria 
 
 in water 282 
 
 Influence of time on efficiency of charcoal 
 
 filters 206 
 
 Influence of variable rates of filtration . . . 215 
 Influent and effluent regulators . . . 209, 214, 247 
 Inoculation of guinea pigs with b. coli com- 
 
 munis 61 
 
 Intermittent sand filters 203 
 
 Iron in ground waters, reduction of .... 203 
 
 Jacksonville, Fla., ground water supply . . 82 
 Jersey City, N.J., consumption of water . . 34 
 Jersey City, N.J., reduction of typhoid fever, 35 
 Jersey City, N. J., typhoid fever rates, 32, 33, 34, 35 
 Jersey City, N.J., water supply of .... 32 
 
 Jewell mechanical filters 186, 195 
 
 Judicial decisions on quality of water supply, 109,287 
 
 Kent Works, London, ground water supply, 83 
 
 Keyport, N.J., mechanical filters 264 
 
 Koch's Comma Bacilliis 276, 280 
 
 'Kronke method of iron reduction 205 
 
 Lake Miiggel water 140 
 
 Lake water, filtration of 20 
 
 Lake Zurich as a source of water supply . . 20 
 
 Lausen, typhoid fever epidemic 91 
 
 Lawrence, Mass., city filter 166 
 
 Lawrence, Mass., epidemic of typhoid ... 98 
 Lawrence, Mass., reduction of typhoid fever in, 36 
 Lawrence, Mass , storage reservoir .... 156 
 Lawrence, Mass.,watersupplyandtyphoidrates, 36 
 Legal Liability of Cities and Water Com- 
 panies for Damages by Sewage-Pol- 
 luted Water 287 
 
 Leipsic, ground water supply 83, 84 
 
 Lime, reduction of albuminoid ammonia by . 118 
 Lime, reduction of bacteria by . . 112, 114, 117 
 
 Lime, reduction of hardness by 117 
 
 Lime, reduction of total solids by 118 
 
 Lime treatment of London water 118 
 
 Lime treatment, cost per million U. S. gallons, 119 
 Limiting the head on filters, Berlin . . 136, 234 
 
 Limpidity and purity of water 12 
 
 Liquefying bacteria 274 
 
 Local causes of typhoid fever 39 
 
 Location of cities, ruling factor 90 
 
 London, a city of the second class 76 
 
 London and Philadelphia, water consumption 
 
 compared 82 
 
 London, treatment of water by lime . . 118, 119 
 Long Branch, N.J., mechanical filter . . . 195 
 Longevity of the typhoid bacillus in water . 17, 42 
 Lorain, Ohio, mechanical filters 195 
 
 PAGE 
 
 Loss by typhoid fever in large cities of Europe 
 
 and Australia 270, 271 
 
 Loss by typhoid fever in large cities of United 
 
 States and Canada 109, 268, 269 
 
 Loss of salts and minerals in water by distilla- 
 tion 85 
 
 Lowell, Mass., epidemic of typhoid .... 98 
 
 Lowell, Mass., filter-bed 175 
 
 Lowell, Mass., ground water supply . . .37, 82 
 Lowell, Mass., reduction of typhoid fever . . 36 
 Lowell, Mass., water supply and typhoid rates, 36 
 Lyons, France, filter gallery 179 
 
 Mager sand-scraping apparatus .... 224 
 
 Management of filters 83 
 
 Management of filters, Berlin .... 234, 235 
 Manchester, Eng., water supply of .... 5 
 Malvoz' experiments with b. coli communis . 64 
 Marston Lake as a source of water supply . . 106 
 
 Af. Aurantiacus 278 
 
 Maximum rate of filtration, London .... 259 
 Maximum yield of filters, Hamburg .... 2lIO 
 
 M. Candicans 278 
 
 M. Carneus 278 
 
 Measurement of efficiency of filters .... 155 
 
 Measurement of the bacteria 278 
 
 Mechanical filters in American cities .... 193 
 
 Mechanical filters, Lorain, Ohio 266 
 
 Mechanical niters, Philadelphia, Pa 197 
 
 Mechanical Filters, Rate of Filtration . 184 
 Mechanical filters, Somerville and Raritan,N.J. 194 
 
 Mechanical sand filters 158, 184 
 
 Memphis, Tenn., ground water supply ... 82 
 Merrimac River as a carrier of typhoid infec- 
 tion 98 
 
 Method of bacterial examination, Miquel . . 243 
 Method of operating Morison filter .... 186 
 
 Micrococci, The - . . 275 
 
 Middletown, Conn., epidemic of typhoid fever, 106 
 Milk as a carrier of typhoid infection . . . 107 
 Milk, infection of, by typhoid-tainted water . 108 
 Minimum rate of filtration, London .... 259 
 Minute amount of organic matter in distilled 
 
 water 17 
 
 Morison mechanical filter 172, 184 
 
 Morison mechanical filter, bacterial efficiency 
 
 of 172 
 
 Motility of the bacteria 276, 277 
 
 Motility of b. typhosus 276 
 
 Mountain sources of water supply . . 37, 81, 90 
 
 Munich, consumption of beer in 80 
 
 Munich, consumption of water in ..... 80 
 Munich, decline of typhoid fever in .... 67 
 
 Munich, source of water supply 16 
 
 Mycoprotein 272 
 
 Natural and alum nitration, relative effi- 
 ciency of 173,198 
 
 Natural filtration through the drift . . 138, 139 
 
INDEX. 
 
 301 
 
 PAGE 
 
 Natural purification of water in the drift . . 22 
 
 Natural sand nitration 139, 158 
 
 Natural sources of pure water not generally 
 
 available 81 
 
 Nencki on chemical composition of the bac- 
 teria 272 
 
 Nitrates and nitrites 273 
 
 Nitrifying bacteria 273 
 
 Nitrifying bacteria in sand-beds of filters, 133, 136 
 
 Non-liquefying bacteria 274 
 
 Nostoc .87 
 
 . 10 
 . 210 
 . 31 
 . 87 
 . 112 
 
 Object of water purification . . 
 
 Objections to filters of large area . 
 
 Objections to sewage-polluted water . 
 
 Odors and tastes in stored waters . . 
 
 Ohio River water, lime treatment of . 
 
 Ohio River water, sedimentation of . . . .111 
 
 Ohio River water, test of, in Parietti solution . 50 
 
 Open filters 158, 210 
 
 Open filters, Altona 252 
 
 Open or closed filters for Cincinnati .... 251 
 Operation of London filters, 144, 145, 146, 147, 148 
 
 Organic matter on watersheds 14 
 
 Oscillaria, 87 
 
 Oysters as a cause of typhoid infection . . . 106 
 
 Parasitic bacteria 273 
 
 Parietti solution, test of Ohio River water with, 
 
 49,50 
 
 Paring of sand-bed 133 
 
 Paris, France, Anderson Purifiers 182 
 
 Paris, France, Storage Reservoirs .... 156 
 Pathogenic bacteria . . . 2, 15, 103, 277, 278, 280 
 Pathogenic bacteria in plate cultures . . . .277 
 
 Pathogenic organisms in water 29, 280 
 
 Pedesis 277 
 
 Percentage of typhoid mortality 66 
 
 Percentage of filtered water available for con- 
 sumption 263, 264 
 
 Percentage of filtered water required for wash- 
 ing sand 263, 264 
 
 Period of operation of filters . . . 161, 228, 261 
 Period of operation of filters, Berlin . . 261, 262 
 Period of operation of filters, Hamburg, 
 
 216, 217, 220, 261 
 
 Period of operation of filter, Lawrence, Mass., 262 
 Period of operation of filters, London . . . 262 
 Period of operation of filters, Lowell, Mass. . 176 
 Period of operation of filters, Providence, R.I., 
 
 171, 172, 262 
 Period of operation of filters, Zurich . . . .261 
 
 Perth, Scotland, filter gallery 179 
 
 Petroleum in drinking-waters 86 
 
 Pettenkofer's theory of typhoid fever . . 66, 69 
 Pequannock River as a source of water supply, 33 
 Philadelphia and London, water consumption 
 
 compared 82 
 
 Piefke method for reduction of iron in ground 
 waters . . 205 
 
 PAGE 
 
 Plymouth, Pa., epidemic of typhoid fever . . 94 
 
 Poisonous minerals in water 86 
 
 Pollution of lakes 20 
 
 Pollution of shallow wells 22 
 
 Pollution of water by fertilizers 23 
 
 Population of the large cities of Australia, 
 
 Europe, and United States, 268, 269, 270, 271 
 Potash alum, treatment of water by . . . .112 
 Poughkeepsie, settling-basins and filters . . 178 
 Precautions with underground cisterns ... 86 
 Production of gas by b. coli communis and 
 
 b. lac t is aerogenes 57 
 
 Products of bacterial action 273 
 
 Propagation of typhoid fever by drinking- 
 water 65 
 
 Proportion of water for domestic uses ... 78 
 Proposed rate of filtration, Denver, Col. . . 260 
 
 Protection of River Thames 17 
 
 Protection of watersheds 21, 76, 81 
 
 Providence experimental filter tests .... 170 
 
 Ptomains, The 3, 4, 278 
 
 Ptomains in water 4, 278 
 
 Public improvements and politics . . . 126, 254 
 
 Pure and impure water 2 
 
 Pure and Purified Water 81 
 
 Pure water not found in nature 3 
 
 Putrefactive bacteria 2, 103 
 
 Putrid gases as predisposing causes of typhoid 
 fever 61 
 
 Quality of water from filter galleries . . . 179 
 Quincy, 111., storage reservoir 157 
 
 Rainfall and Typhoid Fever 104 
 
 Rare occurrence of sources of water supply at 
 
 high elevation 16, 90 
 
 Raritan and Somerville, N.J., mechanical fil- 
 ters of 194 
 
 Rate of delivery of filters 161 
 
 Rates of filtration . . 26, 161, 165, 168, 259, 260 
 
 Rate of filtration, Amsterdam 182 
 
 Rate of filtration, Ashland, Wis 260 
 
 Rate of filtration, Berlin 232 
 
 Rate of filtration, Cincinnati, proposed for, 247, 249 
 Rates of filtration, Fischer filter, Worms . . 240 
 Rate of filtration, The Hague ...... 181 
 
 Rate of filtration, Hamburg 217 
 
 Rate of filtration, Hudson, N.Y 178 
 
 Rate of filtration, Lawrence, Mass. . . 143, 169 
 
 Rate of filtration, London 183 
 
 Rate of filtration, Lowell, Mass 176 
 
 Rate of.filtration, mechanical filters .... 184 
 Rate of filtration, M orison filter . . . 188, 190 
 
 Rate of filtration, Paris 182 
 
 Rate of filtration, Poughkeepsie, N.Y. . . .178 
 
 Rate of filtration, Providence, R.I 171 
 
 Rate of filtration, Rotterdam 180 
 
 Rate of filtration, Tacoma 260 
 
 Rate of filtration, Zurich 143, 183 
 
302 
 
 INDEX. 
 
 PAGE 
 
 Rate of growth of bacteria 276 
 
 Rate of liquefaction of gelatin 172 
 
 Rate of percolation into filter galleries . . .179 
 Rate of percolation through the drift .... 26 
 Reading, Mass., mechanical filters .... 207 
 Recovery of heat from distilled water . . . 122 
 Reduction of albuminoid ammonia by lime 
 
 treatment of water 118 
 
 Reduction of bacteria by alum treatment of 
 
 water 112, 115 
 
 Reduction of bacteria by filtration .... 165 
 Reduction of bacteria by Lawrence filter, 153, 169 
 Reduction of bacteria by lime treatment of 
 
 water 112, 113, 114, 117 
 
 Reduction of bacteria by sedimentation . Ill, 112 
 Reduction of bacteria by sterilized clay . . .113 
 Reduction of color by mechanical filtration . 191 
 Reduction of hardness by lime treatment of 
 
 water 117 
 
 Reduction of iron in ground waters .... 203 
 Reduction of organic matter by subsidence . 117 
 Reduction of sand-bed by scraping . . 182, 235 
 Reduction of silt by subsidence . . 115, 116, 246 
 Reduction of suspended matter by subsidence, 246 
 Reduction of total solids by lime treatment of 
 
 water 118 
 
 Reduction of total solids by mechanical filters, 194 
 Reduction of typhoid rates, Hamburg ... 77 
 Reduction of typhoid rates, Jersey City, N.J. . 34 
 Reduction of typhoid rates, Lawrence, Mass., 
 
 36,37 
 
 Reduction of typhoid rates, Lowell, Mass. . 36 
 Reduction of typhoid rates, Newark, N.J. .33 
 
 Refilling of filters 219, 234 
 
 Regulating valves and weirs . . . 158, 209, 215 
 Relative dimensions of Hamburg filters . . 210 
 Relative efficiency of filters with and without 
 
 alum 173, 198 
 
 Renewal of sand-bed 161, 171 
 
 Reservoirs for filtered water 156 
 
 Reservoirs for ground waters 86 
 
 Restoration to service of sand filters .... 133 
 
 Rivers as carriers of sewage 17 
 
 Rivers as sources of water supply . . . . 15, 16 
 Rivers constitute the principal sources of water 
 
 supply 16 
 
 Rotterdam, filters of 180 
 
 Royal Commission on Metropolitan water sup- 
 
 Ply 4 
 
 Salts and gases in solution in natural waters, 121 
 Sand and charcoal filters for reduction of iron, 
 
 203,204 
 
 Sands and gravels, selection of 162 
 
 Sand scraped and washed per million gallons, 
 
 Hamburg 263 
 
 Sand scraped and washed per million gallons, 
 
 Lawrence, Mass 262 
 
 Sand filters of London water-works .... 6 
 
 PAGE 
 
 Sand filtration in Europe 131 
 
 Sand-washing machinery 213, 235 
 
 San Francisco, typhoid fever in 102 
 
 Saprophytic bacteria 273 
 
 Sarctna, The 275 
 
 Sault Ste. Marie, Mich., epidemic of typhoid 
 
 fever 99 
 
 " Schmutzdecke," The 7, 134, 184 
 
 Schroder sand-washing machine, 220, 221, 222, 223 
 Scraping the sand-bed .... 212, 235, 261, 262 
 Scraping the sand-bed under ice ... 158, 224 
 Seasonal distribution of typhoid fever ... 66 
 Seasonal rotation of large bodies of water . . 19 
 Sedimentation and sand filtration in Europe, 
 
 140, 141 
 
 Sedimentation of Chelsea water 112 
 
 Sedimentation of East London water . . . 112 
 Sedimentation of Grand Junction water . .112 
 
 Sedimentation of Hamburg water 208 
 
 Sedimentation of Lake Linthrathen water . . Ill 
 Sedimentation of Lambeth water ...... 112 
 
 Sedimentation of Ohio River water, 
 
 111, 115, 116,246 
 Sedimentation of Polluted "Waters . . 110 
 
 Sedimentation of Thames water 112 
 
 Sedimentation of West Middlesex water . . 112 
 Sedimentation, time allowed for, in various 
 
 water-works of Europe 141 
 
 Seine, The, water of 102 
 
 Selection of sands and gravels 162 
 
 Self-purification of sewage-polluted streams . 20 
 Separate services for sterilized water . . . 125 
 Settling-basins and filters, Hamburg .... 208 
 
 Sewage infection of oysters 107 
 
 Sewage in river and lake waters 15 
 
 Sewage pollution of drinking-water .... 98 
 Sewage pollution of navigable streams and 
 
 lakes 98 
 
 Sewage-polluted water and disease .... 8 
 Sources of water supply in driven wells . . 82 
 Species of bacteria in water .... 132, 279, 280 
 Spirilla among the water bacteria . . 276, 279, 280 
 
 Spirilla or vibrios, The 276 
 
 Spore-bearing bacteria 279, 286 
 
 Spring Water, N.Y., epidemic of typhoid 
 
 fever 96 
 
 Staining of bacteria 281 
 
 Stamford, Conn., epidemic of typhoid fever . 107 
 
 Standard of filtrate 190, 191 
 
 Standard of food quality 29 
 
 Standard of successful filtration . . . . . 24 
 
 Standard of water quality 3 
 
 Staphylococci, J^he 275 
 
 Staphylococcits pyogenes aureus 278 
 
 Starting a filter in service 136 
 
 Stas-Otto method for reduction of toxalbumens, 62 
 
 Sterilization of articles of diet 38 
 
 Sterilization of Drinking- Water ... 120 
 Sterilization of filter sand 165 
 
INDEX. 
 
 303 
 
 PAGE 
 
 Sterilization of water in U. S. Navy . . 85, 128 
 Sterilization of water at World's Fair . . . 128 
 
 Sterilized water for dietetic uses 85 
 
 Sterilized water on ocean steamships . . 85, 130 
 Sterilizing large quantities of water, cost of . 121 
 St. Louis, Mo., epidemic of typhoid .... 99 
 
 Stone disk and tube niters 47 
 
 Straining action of sand niters 132 
 
 Streptococci, The 275 
 
 Storage and distributing reservoirs for filtered 
 
 water 156, 157 
 
 Storage of filtered water, Lawrence, Mass. . 156 
 
 Storage of filtered water, London 157 
 
 Storage of filtered water, Paris 156 
 
 Storage of filtered water, Quincy, 111. . . . 157 
 
 Storage of ground waters 86 
 
 Storage of surface waters 87 
 
 Storage of water after filtration . . . 156, 157 
 Storage of water from Anderson Purifiers . . 182 
 Source of water supply, Dresden . . . . 75, 88 
 Source of water supply, Hamburg . 77, 88, 208 
 Source of water supply, Lawrence, Mass. . . 88 
 Source of water supply, Liverpool .... 88 
 Source of water supply, Munich . . . . 16, 88 
 
 Sources of impurities in water 2 
 
 Sources of infectious disease 37 
 
 Sources of naturally pure water 81 
 
 Sources of water supply 14 
 
 Sources of water supply, Berlin . . . . 75, 88 
 Sources of water supply, cities of Holland . . 75 
 Sources of water supply, Copenhagen ... 88 
 Sources of water supply, driven wells ... 82 
 
 Sources of water supply, London 88 
 
 Sources of water supply, Paris 88 
 
 Sources of water supply, Stockholm .... 88 
 Sources of water supply, Vienna ... 5, 16, 88 
 South Bend, Ind., ground water supply ... 82 
 South Platte River as a source of water sup- 
 ply 106 
 
 Submerged sand filter, Zurich 96 
 
 Subsidence of organic matter in water . 21, 140 
 
 Subsidence of polluted waters 140 
 
 Subsidence rate of, in Ohio River water . . 116 
 
 Sulphate of alumina analysis 189 
 
 Sulphate of alumina and free flow 188 
 
 Sulphuric acid in filtrate 191 
 
 Tastes and odors in stored waters .... 87 
 Temperature of Hamburg water . . . 210, 220 
 
 Test for alum in filtrate 190 
 
 Test of Ohio River water with Parietti solution, 50 
 
 Tests of water quality 31 
 
 Tetrads, The 275 
 
 The Hague, filters of 181 
 
 Theory of alum and sand filtration .... 188 
 
 Theory of liquefaction of gelatin 274 
 
 Theory of sand filtration ... .... 132 
 
 The Typhoid Bacillus and Typhoid Fever, 56 
 Time required to scrape filter-beds .... 226 
 
 Time required to wash sand-bed 191 
 
 Toxalbumens, Stas-Otto method, reduction by, 62 
 
 Toxic substances in bouillon 3 
 
 Toxin of diphtheria 278 
 
 Toulouse, France, filter gallery 179 
 
 Transmission of typhoid by Furlen Brook . . 92 
 Transmission of typhoid by River Limmat . . 96 
 Transmission of typhoid by Mississippi River, 100 
 Transmission of typhoid by well water ... 93 
 
 Treatment of polluted waters 24 
 
 Treatment of sewage 24 
 
 Typhoid bacillus, 
 
 9, 11, 57, 275, 276, 277, 278, 280, 281, 282, 283 
 Typhoid bacillus, identification of, 56, 64, 278, 288 
 
 Typhoid bacillus, longevity of 17 
 
 Typhoid bacillus, The, in water supply . . 65, 67 
 Typhoid death rates of cities using filtered river 
 
 water 23 
 
 Typhoid death rates of cities using lake water, 22 
 Typhoid death rates of cities using river water, 23 
 
 Typhoid fever and rainfall 105 
 
 Typhoid fever as a measure of city sanitation, 69 
 Typhoid fever as a water-carried disease . . 108 
 Typhoid fever as an autumn disease .... 66 
 Typhoid fever as an index of water quality, 22, 70 
 
 Typhoid fever, Denver, Col 104 
 
 Typhoid fever in Hamburg and Altona ... 68 
 
 Typhoid fever in San Francisco 102 
 
 Typhoid fever, loss by, in large cities of United 
 
 States 109 
 
 Typhoid fever rates, Berlin and Rotterdam . 139 
 Typhoid fever rates, Jersey City, N.J. ... 33 
 Typhoid fever rates, Lawrence, Mass. ... 36 
 
 Typhoid fever rates, Lowell, Mass 36 
 
 Typhoid fever rates, Newark, N.J 33 
 
 Typhoid fever rates, Vienna and Munich . 138 
 Typhoid fever statistics from the larger cities 
 
 of Australia, Canada, Europe, and United 
 
 States ....... 268,269,270,271 
 
 Typhotoxin 62 
 
 Types of mechanical filters 185 
 
 Types of Sand Filters 158 
 
 Tyrotoxicon 278 
 
 Ubiquitous nature of decomposing or- 
 ganic matter 16 
 
 Undecomposed alum in subsided water . . . 115 
 Uniformity coefficient of mixed sizes of sand, 
 
 164, 165, 172 
 Uniformity of discharge from filters .... 232 
 
 Unpotable water, general use of 120 
 
 Use and waste of water in German cities . . 82 
 Use by cities of water from driven wells . 82, 84 
 Use by cities of water from mechanical filters . 89 
 Use of coagulant with mechanical filters . . 184 
 Use of filtered water for washing sand . . . 214 
 Use of water from public mains in large cities, 89 
 
 Vanne water (Paris) 83, 88, 102 
 
304 
 
 INDEX. 
 
 PAGE 
 
 Vanne water, bacterial contents of .... 243 
 Variable consumption of alum with mechanical 
 
 niters 201 
 
 Variation in counts of bacteria 40 
 
 Variation of head on filter-beds 234 
 
 Ventilation of closed niters 235 
 
 Vienna, decline oi typhoid fever in . . . G7 
 Vienna, sources of water supply .... 5, 16 
 Vital products of putrefactive bacteria . . . 278 
 
 "Warren, Ohio, judicial decision on water 
 
 supply 109 
 
 Washing and storage of sand 235 
 
 Washing filter sand and gravel . . . . 161, 213 
 Washing filter sand, water required . . 2G3, 284 
 
 Water an essential of health 1,90 
 
 Water and food as distributers of typhoid in- 
 fection 68 
 
 Water, distrust abroad of natural sources . . 29 
 
 Water from limestone regions 14 
 
 Water from the Danube 5 
 
 Water from the River Seine . . . 102, 242, 243 
 
 Water from the sand dunes 79 
 
 Water from the Vanne Springs .... 102, 243 
 
 Water, percentage of body weight 1 
 
 Water, percenfage of, in the circulation ... 1 
 Water required by Schroder sand washers . . 223 
 Water required for washing sand, Asbury Park, 
 
 N.J 264 
 
 Water required tor washing sand, Berlin . . 263 
 Water required for washing sand, Hamburg, 
 
 263,264 
 Water required f orwashing sand, Keyport , N . J ., 264 
 
 PAGE 
 Water required for washing sand, Long Branch, 
 
 N.J 264 
 
 Water space in sand-bed 164, 232 
 
 Water storage after filtration 156 
 
 Water supply, evidence of b. typhosns i:i . 60, 288 
 Water supply in mountain sources .... 37 
 Water supply in the Mangfall Valley ... 16 
 
 Water supply in the Schneeberg 16 
 
 Water supply of Hamburg . . .... 77 
 
 Water supply of Jersey City, N.J . . . 32, 33 
 
 Water supply of Lawrence, Mass 36 
 
 Water supply of Liverpool 81 
 
 Water supply of Lowell, Mass 36, 37 
 
 Water supply of Manchester, Eng. . . . 5, 81 
 
 Water supply of Newark, N.J 32, 33 
 
 Water supply of New York and Edinburgh . 76 
 Water supply of San Francisco ... .28 
 
 Water supply of Vienna 5, 16, 88 
 
 Water the cause of continuous typhoid fever 
 
 rates . . . . 39 
 
 Water transmission of infectious disease . . 9 
 Welsh sources of water supply for London, 84, 118 
 Winter temperatures, Cincinnati, etc. . . . 251 
 Worcester, Eng., purification works . . . 244 
 Wurtz, milk-sugar, litmus agar 103 
 
 Yaryan multiple effect sterilizer, 125, 127, 130 
 
 Yield of filters after scraping sand . . . 227, 228 
 Yield of wells in Germany 84 
 
 Zurich epidemic of typhoid fever ... 96 
 
 Zurich, filters of 183 
 
 Zymotic disease, prophylaxis of 8 
 
LIST OF BOOKS 
 
 ON 
 
 Water Supply and Sanitary Science. 
 
 ADAMS, J. W. Sewers and Drains for Populous Districts. Embracing Rules 
 and Formulas for the dimensions and construction of works of Sanitary 
 Engineers. 8vo, cloth. $2.50. 
 
 BAKER, M. N. Sewerage and Sewage Purification. 18mo, cloth. 50 cents. 
 BROWN, GLENN. Healthy Foundations for Houses. 18mo, boards. Illus- 
 trated. 50 cents. 
 
 CORFIELD, W. H. Water and Water Supply. 18mo, boards. 50 cents. 
 Dwelling Houses; their Sanitary Construction and Arrangements. 18mo, 
 boards. 50 cents. 
 
 FANNING, J. T. A Practical Treatise on Hydraulic and Water-Supply Engi- 
 neering. Relating to the Hydrology, Hydrodynamics, and Practical Con- 
 struction of Water-works in North America. 180 illustrations. 8vo, cloth. 
 Thirteenth Edition, revised, enlarged, and new tables and illustrations added, 
 650 pages. $5.00. 
 
 GERHARD, W. P. Recent Practice in the Sanitary Drainage of Buildings. 
 
 ISmo, boards. 50 cents. 
 
 Disposal of Household Waste. 18mo, boards. 50 cents. 
 House Drainage and Sanitary Plumbing. 18mo, boards, lllus. 50 cents. 
 
 HILL, JOHN W. Water Analyses and their Interpretation. 12mo, cl. (In press .) 
 
 KIRKWOOD, JAS. P. Report on the Filtration of River Waters for the Sup- 
 ply of Cities, as Practised in Europe, made to the Board of Water Com- 
 missioners of the City of St. Louis. Illustrated by 30 double-plate engrav- 
 ings. 4to, cloth. $7.50. 
 
 MAGUIRE, WM. R. Domestic Sanitary Drainage and Plumbing Lectures on 
 Practical Sanitation. Second Edition. 332 illustrations. 8vo, cloth. $4.00. 
 
 RAFTER, GEO. W., and BAKER, M. N. Sewage Disposal in the United States. 
 Illustrations and folding plates. Second Edition. 8vo, cloth. $6.00. 
 
 RAFTER, GEO. W. The Microscopical Examination of Potable Water. With 
 diagrams. 18mo, boards. 50 cents. 
 
 SLATER, J. W. Sewage Treatment, Purification, and Utilization. A Practical 
 Manual for the Use of Corporations, Local Boards, Medical Officers of 
 Health, Inspectors of Nuisances, Chemists, Manufacturers, Riparian Own- 
 ers, Engineers, and Rate-payers. 12mo, cloth. $2.25. 
 
 STALEY, CADY, and PIERSON, GEO. S. The Separate System of Sewer- 
 age : Its Theory and Construction. 8vo, cloth. With maps, plates, and 
 illustrations. Second Edition. $3.00. 
 
 TIDY, C. M. The Treatment of Sewage. 18mo, boards. 50 cents. 
 
 VARONA, A. de. Sewer Gases : their Nature and Origin. 18mo, boards. 50 cents. 
 
 WARING, G. E. Sewerage and Land Drainage. Third Edition. 4to, cloth. 
 
 Illustrated, colored plates. $6.00. 
 
 Sanitary Condition of City and Country Dwelling Houses. 18mo, bds. 50 cts. 
 The Sanitary Drainage of Houses and Towns. 12mo, cloth. $2.00. 
 How to Drain a House. Practical Information for Householders. 12mo, cloth. 
 
 $1.25. 
 
 Modern Methods of Sewage Disposal for Towns, Public Institutions, and Iso- 
 lated Houses. $2.00. 
 
 D. VAN NOSTRAND COMPANY, PUBLISHERS, 
 
 23 MURRAY and 27 WARREN STS., NEW YORK. 
 Copies sent by mail on receipt of Price. 
 
THIRTEENTH EDITION. 
 
 I Vol., octavo, 644 pp., 200 Illustrations, fine Cloth Binding, $5.00 
 
 A PRACTICAL TREATISE 
 
 WATER-SUPPLY ENGINEERING 
 
 RELATING TO THE 
 
 HYDROLOGY, HYDRODYNAMICS, AND PRACTICAL CONSTRUCTION 
 OF WATER-WORKS IN NORTH AMERICA, 
 
 WITH NUMEROUS 
 
 TABLES AND ILLUSTRATIONS. 
 
 BY 
 
 J. T. FANNING, CE, 
 
 Member of tlie A merican Society of Civil Engineers. 
 
 Thirteenth Edition, Revised, Enlarged, and New Tables and Illustrations added. 
 
 CONTENTS. 
 
 SECTION I. Collection and Storage of Water, and its Impurities. 
 CHAPTER I. Introductory. CHAP. II. Quantity of Water required. CHAP. III. 
 Rainfall. CHAP. IV. Flow of Streams. CHAP. V. Storage and Evaporation 
 of Water. CHAP. VI. Supplying Capacity of Watersheds. CHAP. VII. 
 Springs and Wells. CHAP. VIII. Impurities of Water. CHAP. IX. Well, 
 Spring, Lake, and River Supplies. 
 
 SECTION II. Flow of Water through Sluices, Pipes, and Channels. 
 CHAPTER X. Weight, Pressure, and Motion of Water. CHAP. XI. Flow of Water 
 through Orifices. CHAP. XII. Flow of Water through Short Tubes. CHAP. 
 XIII. Flow of Water through Pipes under Pressure. CHAP. XIV. Measures 
 of Weirs and Weir Gauging. CHAP. XV. Flow of Water in Open Channels. 
 
 SECTION III. Practical Construction of Water-Works. 
 
 CHAPTER XVI. Reservoir Embankments and Chambers. CHAP. XVII. Open Ca- 
 nals. CHAP. XVIII. Waste Weirs. CHAP. XIX. Partitions and Retaining 
 Walls. CHAP. XX. Masonry Conduits. CHAP. XXI. Mains and Distribution 
 Pipes. CHAP. XXII. Distribution Systems and Appendages. CHAP. XXIII. 
 Clarification of Water. CHAP. XXIV. Pumping of Water. CHAP. XXV. Tank 
 Stand Pipes. CHAP. XXVI. Systems of Water Supply. 
 
 APPENDIX. Miscellaneous Memoranda. 
 
 D. VAN NOSTRAND COMPANY, Publishers, 
 
 23 MURRAY and 27 WARREN STS., NEW YORK. 
 *** Copies sent postpaid on receipt of price. 
 
A STANDARD WORK. 
 One Volume, quarto, cloth, 30 plates ...... Price, $7.50 
 
 REPORT 
 
 ON THE 
 
 FILTRATION OF RIVER WATERS 
 
 FOR THE SUPPLY OF CITIES, 
 
 AS PRACTISED IN EUROPE. 
 
 MADE TO THE 
 
 Board of Water Commissioners of the City of St, Louis, 
 
 JAMES P. KIRKWOOD, Civil Engineer* 
 
 II. 
 
 III. 
 IV. 
 V. 
 
 VI. 
 
 VII. 
 
 VIII. 
 IX. 
 
 X. 
 
 XI. 
 
 XII. 
 
 XIII. 
 
 XIV. 
 XV. 
 
 BY IPERMIISSICKN" OF 1 THE BOARD. 
 
 ILL USTRA TED B Y THIRTY ENGRA VINCS. 
 
 Report on Filtration. 
 
 London Works, General. 
 
 Chelsea Water Works and Filters. 
 
 Lambeth Water Works and Filters. 
 
 Southwark and Vauxhall Water Works 
 and Filters. 
 
 Grand Junction Water Works and 
 Filters. 
 
 West Middlesex Water Works and 
 Filters. 
 
 New River Water Works and Filters. 
 
 East London Water Works and Fil- 
 ters. 
 
 Leicester Water Works and Filters. 
 
 York Water Works and Filters. 
 
 Liverpool Water Works and Filters. 
 
 Edinburgh Water Works and Filters. 
 
 Dublin Water Works and Filters. 
 
 Perth Water Works and Filtering Gal- 
 lery. 
 
 CONTENTS. 
 
 XVI. Berlin Water Works and Filters. 
 XVII. Hamburg Water Works and Res- 
 ervoirs. 
 
 XVIII. Altona Water Works and Filters. 
 XIX. Tours Water Works and Filtering 
 
 Canal. 
 XX. Angers Water Works and Filtering 
 
 Galleries. 
 
 XXI. Nantes Water Works and Filters. 
 XXII. Lyons Water Works and Filtering 
 Galleries. 
 
 XXIII. Toulouse Water Works and Filter- 
 
 ing Galleries. 
 
 XXIV. Marseilles Water Works and Filters. 
 XXV. Genoa Water Works and Filtering 
 
 Galleries. 
 
 XXVI. Leghorn Water Works and Cis- 
 terns. 
 
 XXVII. Wakefield Water Works and Fil- 
 ters. 
 
 APPENDIX. 
 
 Instructions, Tables of Equivalents of Measures, London Pumping Engines Tabulated, Boil- 
 ers of Pumping Engines Tabulated. 
 
 D. VAN NOSTRAND COMPANY, Publishers, 
 
 23 MURRAY and 27 WARREN STS., NEW YORK. 
 
 Copies sent postpaid on receipt of price. 
 
SECOND EDITION. 
 
 One Volume, 8vo, Cloth, Illustrated, 600 Pages. Price, $6.00. 
 
 SEWAGE DISPOSAL 
 
 IN THE UNITED STATES 
 
 By Geo. W. Rafter, M. Am, Soc. C E., and M. N. Baker, Ph. B., Associate 
 Editor, " Engineering News." 
 
 CONTENTS. 
 
 PART I. DISCUSSION OF PRINCIPLES. 
 
 CHAPTER 
 
 I. Preliminary Discussion. 
 II. The Infectious Diseases of Animals. 
 III. On the Pollution of Streams. 
 IV. The Self-Purification of Running Streams 
 and the Rational View in Relation to 
 the Disposal of Sewage by Discharge 
 into Tide- Water. 
 
 V. The Composition of Sewage Muds. 
 Legal Aspects of the Case. 
 Quantity of Sewage and Variation in Rate 
 
 of Flow. 
 
 General Data of Sewage Disposal. 
 Discharge into Tidal or other large Bodies 
 
 of Water. 
 
 On Nitrification and the Nitrifying Organ- 
 ism. 
 
 VI. 
 VII. 
 
 VIII. 
 IX. 
 
 X. 
 
 CHAPTER 
 
 XI. Chemical Precipitation. 
 XII. Broad Irrigation. 
 
 XIII. On Silos and their Use in Sewage Farming. 
 XIV. Intermittent Filtration. 
 XV. Sub-surface Irrigation. 
 XVI. The Disposal of Manufacturing Wastes. 
 XVII. On the Temperature of the Air and of 
 Natural Soils, and its Relation to 
 Sewage Purification by Broad Irriga- 
 tion and Intermittent Filtration. 
 XVIII. On Beggiatoa Alba and its Relation to 
 
 Sewage Effluents. 
 
 XIX. The Effect of the Pollution of Streams 
 by Manufacturing Wastes upon the 
 Life of Fish. 
 XX. Conclusions to Part I. 
 
 CHAPTER 
 XXI. 
 
 XXII. 
 XXIII. 
 
 XXIV. 
 XXV. 
 
 XXVI. 
 
 XXVII. 
 
 XXVIII. 
 
 XXIX. 
 
 XXX. 
 
 XXXI. 
 
 XXXII. 
 
 PART II. -DESCRIPTIONS OF WORKS. 
 
 CHAPTER 
 
 Pail System at Hemlock Lake, New 
 York. 
 
 The Fullerton Avenue Conduit and 
 the Bridgeport Pumping Station, 
 Chicago. 
 
 Chemical Precipitation Plants at Coney 
 Island, Round Lake, White Plains, 
 and Sheepshead Bay, New York. 
 
 Chemical Precipitation and Filtration 
 at East Orange, New Jersey. 
 
 Chemical Precipitation and Mechanical 
 Separation at Long Branch, New 
 Jersey. 
 
 The Mystic Valley Chemical Precipita- 
 tion Works. 
 
 Chemical Precipitation at Worcester, 
 Massachusetts. 
 
 Discharge into Tide-Water and Pro- 
 posed Chemical Precipitation at 
 Providence, Rhode Island. 
 
 Broad Irrigation at the State Hospital 
 for the Insane, Worcester, Mass. 
 
 Broad Irrigation and Intermittent Fil- 
 tration at Pullman, Illinois. 
 
 Broad Irrigation at the Massachusetts 
 Reformatory, Concord. 
 
 Broad Irrigation at the Rhode Island 
 State Institutions. 
 
 XXXIII. Intermittent Filtration and Broad Ir- 
 
 rigation at South Framingham, 
 Massachusetts. 
 
 XXXIV. Intermittent Filtration at Medfield, 
 
 Massachusetts. 
 
 XXXV. Intermittent Filtration and Broad Ir- 
 rigation at the London, Ontario, 
 Hospital for the Insane. 
 
 XXXVI. Chemical Precipitation and Intermit- 
 tent Filtration at the Rochester, 
 Minnesota, Hospital for the Insane. 
 XXXVII. Intermittent Filtration at Marlbor- 
 
 ough, Massachusetts. 
 
 XXXVIII. Intermittent Filtration at the Massa- 
 chusetts School for the Feeble- 
 Minded. 
 
 XXXIX. Sub-surface Irrigation at Lawrence- 
 
 ville, New Jersey, School for Boys. 
 
 XL. Intermittent Filtration at Gardner, 
 
 Massachusetts. 
 XLI. Intermittent Filtration at Summit, 
 
 New Jersey. 
 
 XLII. Land Disposal at Hastings, Nebraska. 
 XLIII. Surface Irrigation at Wayne, Penn- 
 sylvania. 
 XLIV. The Use of Sewage for Irrigation in 
 
 the West. 
 XLV. Miscellaneous Plants. 
 
 D. VAN NOSTRAND COMPANY, Publishers, 
 
 23 MURRAY and 27 WARREN STS., NEW YORK. 
 *** Copies sent by mail on receipt of price. 
 
THIS BOOK IS DUE ON THE LAST DATE 
 STAMPED BELOW 
 
 AN INITIAL FINE OF 25 CENTS 
 
 WILL BE ASSESSED FOR FAILURE TO RETURN 
 THIS BOOK ON THE DATE DUE. THE PENALTY 
 WILL INCREASE TO SO CENTS ON THE FOURTH 
 DAY AND TO $I.OO ON THE SEVENTH DAY 
 OVERDUE. 
 
 OCT 171930 
 
 REC'D 
 
 LOAN DE 
 
 1977 
 
 I0 t IS? 
 A/97 
 
 ? 
 
 
 
 MAR 
 
 T. 
 
 LD 21-95m-7,'37