UC-NRLF 1906 ! ^ ^urO^ lA/Ofc ^_ LIBRARY OF THE UNIVERSITY OF CALIFORNIA. Class MONOGRAPHS ON APPLIED ELECTROCHEMISTRY EDITED BY VIKTOR ENGELHARDT, Chief Engineer and Chemist of the Siemens & Halske Co., Limited, Vienna. WITH THE COOPERATION OF Dr. E. Abel, Chemist for the Siemens & Halske Co., Ltd., Vienna. Dr. P. Askenasy, technical manager of the Accumulator plant, Liesing. H. Becker, editor of " L'Industrie electro-chimique, " Paris. Dr. W. Borchers, Professor at the Technical High School, Aachen. Sh. Cowper-Coles, Editor of " The Electrochemist and Metallurgist, " London. Dr. F. Dieffenbach, Professor at the Technical High School, Darm- stadt. Dr. G. Erlwein, Chief Chemist of the Siemens & Halske Co., Ltd., Berlin. H. Friberg, Engineer of the Siemens & Halske Co., Ltd., Berlin. H. Gall, Director of the Socie"te" d'Electrochimie, Paris. F. E. Giinther, Mining Engineer, Aachen. Dr. F. Haber, Professor at the Technical High School, Karlsruhe. Dr. C. Haussermann, Professor at the Technical High School, Stutt- gart. Dr. R. Hammerschmidt, Electrochemist, Charlottenburg. Dr. G. Hausdorff, Official Chemist, Essen. Dr. K. Kellner, General-director, Vienna. A. Krakau, Professor at the Electrotechnical Institute, St. Petersburg. Dr. H. Landolt, Director of the Society of Electrochemical Industry, Turgi. Dr. M. Le Blanc, Professor at the Technical High School, Karlsruhe. C. Liebenow, Engineer, Berlin. Dr. R. Lorenz, Professor at the Swiss Polytechnick, Zurich. Dr. R. Lucion, Director of Solvay & Co., Briissel. A. Minet, Editor of " L'Electrochimfe, " Paris. A. Nettel, Engineer, Berlin. H. Nissenson, Director of the Stolberg & Westfalen Co., Ltd., Stol- berg. Dr. F. Peters, Instructor at the Royal Mining Academy, Berlin. Dr. W. Pfanhauser, Manufacturer, Vienna. Registered Chemist Dr. O. Prelinger, Chemist for the Siemens & Halske Co., Ltd., Vienna. Dr. Th. Zettel, Chief Chemist of Brown-Boveri & Co., Baden. And other experts. MONOGRAPHS ON APPLIED ELECTROCHEMISTRY. VOL I. THE ELECTROLYSIS OF WATER PROCESSES AND APPLICATIONS BY VIKTOR ENGELHARDT rr CHIEF ENGINEER AND CHEMIST OF THE SIEMENS & HALSKE Co., LIMITED, VIENNA. AUTHORIZED ENGLISH TRANSLATION BY JOSEPH W. RICHARDS, M.A., A.C., PH.D., PRESIDENT OF THE AMERICAN ELECTROCHEMICAL SOCIETY. PROFESSOR OF METALLURGY AT LEHIGH UNIVERSITY. With 90 Figures and 15 Tables in the Text. OF THE ^V UNIVERSITY j EASTON, PA. : PUBLISHED BY THE CHEMICAL PUBLISHING COMPANY. 1904. COPYRIGHT, 1904, BY THE CHEMICAL PUBLISHING Co. OF THE 'DIVERSITY Vor *AUFOW AUTHOR'S PREFACE TO GERMAN EDITION. Corresponding with the rapid growth which the application of electrochemical processes have shown in practice, the literature of this field has grown remarkably in the last ten years of the past century. This literature has, however, in spite of the existence of present German and other treatises and the regularly appearing periodicals, yet exclusively re- mained a simple current report. Those- works which have appeared, which in a concise form treat of the larger field' of applied electrochemistry, are for the most part primarily in- tended for students and therefore have a general character. Special treatises on applied electrochemistry which will treat of their subjects exhaustively in all directions are, with the very few exceptions, not in existence. This deficiency is, however, easily explainable if we con- sider more closely the way in which electrochemical processes reach commercial applications. We can, on the whole, if we leave out of consideration the electrochemical sources of the current, which form a group of themselves, divide these electrochemical processes into two principal classes. One series of processes which we may designate as electro- chemical installation-processes, have as their object, with- out forming a branch of manufacture themselves, to introduce technical improvements or economies in the existing indus- tries and processes. What appears on such processes in literature is only in rarest cases published by the industries making use of the processes themselves ; but in the main by the patentees of the process in question, who are in most cases, identical with the manufacturers of the machinery and apparatus necessary for the practice of the process. Since it is in the interest of the latter to erect as many plants as possible, the literature must be exploited for advertising purposes. The description iv AUTHOR'S PREFACE. given of the form of apparatus is usually only diagrammatic, and often has the object of deceiving those who should hope to work in the same direction. As an example of such elec trochemical installation we may quote the electrolytic puri- fication of beet-sugar solution, the so-called electrical bleach- ing, many applications of ozone, the electrolytic decomposition of water, some metallurgical refining processes, etc. The second large group are the real electrochemical manu- facturing processes like the chlorine and alkali industry, the manufacture of chlorates, carbides, most of the metallurgical applications of electrochemistry and the like. In these pro- cesses the withholding of published information is greater than in the previous ones. With the exception of the patent specifications, very little is made public. Every practical electrochemist knows how far the principles described in the patent specifications are different from the real applications in practice. Iii the collection of " Monographs of Applied Electrochem- istry," the first volume of which is herewith published, it will be the object to set forth detailed and most authentic reports in the field of applied electrochemistry. These mono- graphs will not be general compilations on the present con- dition of the several fields of applied electrochemistry, but exhaustive special reports, in which the entire historical de- velopment will be set forth, and a good review of the most important patent literature made. An endeavor will further be made to modify, as far as possible, the conservatism of the commercial circles and to give as far as possible commercial data, such as cost of plant and operation, commercial con- ditions and the like. Our co-operators, who on the one side are instructors in elec- trochemistry in the technical colleges, in close connection with the pioneers of electrochemistry, on the other side practical electrochemists in successful commercial work, permit the hope of our success to appear well-founded. Yet we do not like to narrow the boundaries which we will TRANSLATOR'S PREFACE. v observe in the " Monographs of Applied Electrochemistry.'' How much work which was undertaken with a purely theo- retical interest really provides a rich source of valuable mate- rial for industrial applications! How valuable for the prac- tical man would be the exhaustive reports on the special con- ditions of work and production in the different countries in which electrochemistry flourishes! We therefore hope that the collection of treatises whose publication is now begun will be a welcome assistance to our professional men in their work. Surely we will reach this goal if not only the above-named cooperators but all professional men will aid us by commu- nicating their experiences and reporting any modifications or mistakes. For assistance rendered in this direction we therefore offer our best thanks in advance. VIKTOR ENGELHARDT. Vienna, January, 1902. TRANSLATOR'S PREFACE. To any one in the least conversant with the rapid recent growth of applied, as well as theoretical electrochemistry, no apology or explanation will be necessary regarding the time- liness of translating the present series of monographs. A considerable proportion of the English-speaking electrochemists can read the German text, but a still larger proportion can- not, and that the latter class may have access to the thoughts and ideas of these monographs, is the purpose of these trans- lations. The present work is rich in suggestions of ways and means for accomplishing difficult ends, in electrochemical op- erations ; its perusal will be found instructive and stimu- lating to the electrochemist in any line. It is therefore of vi TRANSLATOR'S PREFACE. far greater value than a mere code of operations for producing hydrogen and oxygen, as will be recognized by the reader as soon as he reads into the text. The translator wishes here to acknowledge his indebted- ness to W. S. Landis, E.M., for his considerable assistance in taking down the translation and reading proof-sheets. JOSEPH W. RICHARDS. Metallurgical Laboratory, Lehigh University, Sept. 22, 1903. CONTENTS. I. HISTORICAL REVIEW : Introduction T The Discovery of the Electrolytic Decomposition of Water. i Older Literature 4 II. THE CONSTANTS OF THE ELECTROLYTIC DECOMPOSITION OF WATER : Chemical and Electrochemical Constants 6 A . Oxygen 6 B. Hydrogen 7 C. Detonating Gas 7 Decomposition Voltage 7 Conductivity 8 III. REVIEW OF THE PROCESSES : A. Processes and Apparatus for the Separate Production of Oxygen and Hydrogen 9 (a) With Porous Diaphragms of Non-conducting Material 9 Process of d'Arsonval, 1885 10 Form of Apparatus 10 Process of Latchinoff 1 1 Patent Claim n First Method of Operation 1 1 Later Methods of Operation 12 Drying of the Gases 14 Moderation of Complications 15 Bipolar Connections 15 Electrolysis under Pressure . 15 Plant 17 Output 18 Cost of Operation 18 Practice 18 Process of Ducretet, 1888 19 Form of Apparatus 19 Process of Renard, 1888-1890 19 Output - 20 Form of Apparatus 20 Practice 21 Cost of Plant and Operation 21 Laboratory Apparatus 21 Process of Delmard, 1890 23 Process of Bell, 1893 24 Patent Claim 24 P'orm of Apparatus 25 Practice 27 Vlll CONTENTS. Process of Schmidt, 1899 27 Patent Claim 28 Description 28 Form of Apparatus 30 Output 31 Purity of the Gases 31 Types of Apparatus 32 Arrangement 33 Rules for Operating 33 (a ) Gasometer 33 () Conductors and Water Supply 34 (c) Measuring and Controlling Apparatus 35 (d) Cocks and Joints 35 ( in our case, the electrical energy required must equal the heat of formation. Since the heat of formation of water is 68,400 calories and i volt-coulomb = 4.18 X io 7 ergs = 0.2394 cals, there is, therefore, required for the electrolysis of a gram-molecule of water 68,400 285,714 volt-coulombs. O s r Since in the decomposition of the gram-molecule of water two gram-equivalents of hydrogen are set free at the cathode, which require, according to Faraday's law, 2 X 96540 cou- lombs, the decomposition voltage necessarily must be 285,714 volt-coulombs T- i T = 1.48 or roundly 1.5 volts. 2 X 96,540 coulombs The first theoretical investigations upon, the electromotive force necessary to decompose water were made by Helmholtz, 1 who by different methods obtained for it 1.6447 an ^ I -7^3 volts. LeBlanc determined experimentally the decomposition point of water at 1.67 volts, at which electromotive force the 1 Ges. Abh., Ill, 92 and 267. 8 ELECTROLYSIS OF WATER. permanent passage of the current commences. Glaser re- peated the investigations of LeBlanc in particular to clear up the contradiction that the H-O cell only gives 1.08 volts. Glaser found at this latter voltage a kink in the decomposition curve, but ascribed the same to the separation of doubly charged oxygen ions and explained hereby the contradiction with the gas cell. For further data, reference is made to the orig- inal work. 1 As the main results of a part of this work, Glaser promulgated the law that the decomposition of water, although it can take place primarily, in reality takes place mostly second- arily, especially with moderately strong currents. CONDUCTIVITY. The conductivity of water is, according to the work of Kohlrausch, 2 very small and is for pure distilled water in a column i m. long and i sq. mm. cross-section, at 18 C. 0.04 X io~ 10 (Hg = i). For all practical purposes of electrolytic decomposition of water the latter must first be made conducting. For this purpose acids as well as bases are used. A table in the ap- pendix gives data on resistances and conductivity of electro- lytes used in the technical decomposition of water, that is, of sulphuric acid, caustic alkalies and the alkaline carbonates. M. U. Schoop 3 gives graphical representations of the con- nection between resistances of electrolytes and their concen- tration. III. REVIEW OF THE PROCESSES. Although it would be attractive and, in most respects, justifiable to describe the various apparatus and processes used in the electrolytic decomposition of water in a chrono- logical order, yet the difficulty would arise in doing so, that apparatus of quite various principles would come close to- 1 Zeitschr. fur Blektrochem., 1897-1898, 374. See also Caspar! : tiber Wasserstoffentwicklung, Zeitschr. fur Elektrochem., 1899-1900, 37. 2 Kohlrausch : Ztschr. phys. Chem., 1894, 14, 317. 3 Die industrielle Elektrolyse des Wassers, 1901, 113. REVIEW OF THE PROCESSES. 9 gether. By so doing, clearness of description would suffer. We will, therefore, in the following pages, describe the processes by groups and give a chronological compilation at the close of this section in a tabular form. The division into groups may most advantageously be made in the following way: A. Processes and Apparatus for the Separate Production of Oxygen and Hydrogen. (a) With Porous Diaphragms of Non-conducting Mate- rial. (b) With Complete Non-conducting Partitions. (a) For Instruction and Laboratory Work. (b) For Technical Purposes. (c) With Complete or Perforated Conducting Partitions. B. Processes and Apparatus for the Electroly- sis of Water without Separation of the Gas (Production of Detonating Gas). (a) For Instruction and Laboratory Work. (b) For Technical Purposes. C. Processes for the Simple Evolution of Oxygen. (a) Through Depolarization at the Cathode. (b) By the Precipitation of Metal at the Cathode. A. Processes and Apparatus for the Separate Production of Oxygen and Hydrogen. ^ ^si 2 % Ih Fig. 4. Fig. 5- ing uponyC The vessel a is provided with an overflow, , and is insulated by insulators and wooden stringers. The cathode is formed by the vessel a itself while the piece of sheet-iron b serves as anode. The latter is surrounded by a box made of ebonite strips covered with parchment paper. The anode receives the current by an insulated wire introduced through the overflow underneath the lower edge of the bell. The bell d, likewise of iron, is divided into two spaces, which are in connection with the gas pipes by the T-joints, g and k. 1 Zeitschr. f. Elektrot., 1894, 338, 364 and 382. Electrochem. Zeitschr., 1894-95, 1 06. 14 ELECTROLYSIS OF WATER. Since the bell d is in electrical connection with the cathode vessel, the lower part of the inner chamber of d is lined with hard rubber above the level of the liquid, in order not to con- taminate the oxygen passing into it, by hydrogen which might be thereupon developed. In the apparatus the electrolyte was also a 10-15 per cent, solution of caustic soda as free as possible from carbonates. The apparatus containing not quite 60 litres required 300 amperes. The anode measured 9x5 dm. and had therefore on both sides 90 sq. dm. surface. The outer vessel measured Fig. 6. 50x100x11 cm. The current density at the anode was there- fore 3.5 amperes per sq. dm. The single cells were connected up as shown in Fig. 6. The gases were collected by a series of pipes with connec- tions to each apparatus, and dried before going to the gasom- eter. DRYING OF THE GASES. For a drying chamber a long box with rounded cover was used, with a partition reaching not quite to the bottom, and on both sides of the partition lay a quantity of pumice stone xcr* S or { UNIV UNIVERSIT PROCESS OF LATCHINOFF. 15 soaked with sulphuric acid. The gases entering into one side of the partition were forced to pass through the pumice, under the partition, and up on the other side. The gases are there- fore not only dried but also freed from any particles of alka- line electrolyte carried out with them. MODERATION OF COMPLICATIONS. To avoid irregularities in the electrical decomposition, mer- cury manometers are provided which are connected with one pole of an alarm bell. The second pole of the alarm bell was connected to a platinum wire sealed into the upper part of the manometer tube. When the pressure rises in the apparatus the mercury is forced up the tube of the manometer and when it reaches the platinum wire mentioned, completes the circuit and rings the bell. BIPOLAR CONNECTIONS. Latchinoff was the first to use bipolar electrodes in the electrolytic decomposition of water. For smaller plants as in laboratories, pharmacies, etc., a long box about 2 metres long made of paraffined wood is divided by air-tight electrode plates into a series of chambers. Between the electrodes were placed diaphragms of parchment. All the intermediate partitions acted as double plates since only the two external ones were connected with the source of the electric current. An appa- ratus with about 40 electrodes could therefore be connected di- rectly to a normal lighting circuit. Latchinoff used a current density of about 10 amperes per sq. dm. of electrode surface for such an apparatus. Dr. O. Schmidt, as we shall see later, has used recently this principle of bipolar arrangement of electrodes in a practical apparatus, but Latchinoff was the first to propose it for the de- composition of water. ELECTROLYSIS UNDER PRESSURE. Likewise Latchinoff was the first to propose and carry out in a decomposition apparatus the compression of the gases. 16 ELECTROLYSIS OF WATER. a Proceeding from the assumption that even considerable pres- sure would be without influence upon the output, he arranged the apparatus so that it was connected directly with a steel tube intended to stand a gas pressure of 100-120 atmospheres. The apparatus shown in Fig. 7 consists of a cylindrical steel vessel which is closed by a cover tightly screwed on. The tubu- lar iron electrode b in the middle of the tube is raised upon an insulating stand, g. It is connected with the binding post /by an insulated wire, k. The second binding post /' is di- rectly screwed into the steel cyl- inder, which forms the second pole. The cover carries two pipe connections, one in the middle and the second on one side. At the lower edge of the tubes V |^=lf|E^| = is a conical valve, supported ^S |= =?| /^ upon a cylindrical float, which / inkfll^M^E^L^ jffll/ latter works on a guiding tube not shown in the drawing. An insulated cylinder, whose lower part dips into the electrolyte, serves to prevent the mixing of the gases. The insulating stand g, with its flaring sides opening upwards, serves to direct any gas evolving from the bottom of the steel cylinder towards the sides of this cylinder, so that it rises into the outer space at the top. The apparatus is filled three-fourths full of caustic soda solution. The tube connections e and a are joined with the steel cylinder in which the gases are to be compressed. The oxygen reservoir must only be one-half as large as the hydro- PROCESS OF LATCHINOFF. 1 7 gen reservoir in order that the gas pressure remains the same. It is, however, impossible to avoid small variations in pressure which would lower the level of the liquid in one of the elec- trode spaces, and thereby give opportunity for the formation of detonating gas. This danger is avoided by the use of the floating valves a 1 e' . As soon as the level of the liquid rises higher in one of the compartments the valve, as for instance a 1 ', rises with the rising electrolyte and closes the gas exit, whereupon the gas, being still evolved, reduces the level. As soon as the valve is no longer held in place by the float it falls back and allows the gas in question to pass out. This provides an automatic regulator to the pressure. Unfortunately, there is in the literature of the subject no information as to whether this apparatus was tried on a large scale, and whether or in what manner the increasing absorp- tion of the gases in the electrolyte by increased pressure made itself noticeable. PLANT. LatchinofT gives the following figures for an industrial plant 1 : A 50 H. P. source of energy drives a direct current dynamo furnishing 300 amperes and no volts. The electrolytic plant consists of 44 baths, each one 1.4 metres high, and requires 5x1^ metres of floor space. The number of cells therein mentioned in the journals, namely, 40 in series, must be a mistake, since in this case the production named would be theoretically scarcely possible. Latchinoff used 44 cells which number he also mentions in another place as using with a 100 H. P. plant and no volts working tension. A 50 H. P. plant produces, per hour, 5.5 cubic metres of hydrogen and 2.75 cubic metres of oxygen or together 8.25 cubic metres of gas, or 100 cubic metres of oxygen and 200 cubic metres of hydrogen in thirty-six hours, decomposing 150 litres of water. 1 Elektrochemische Zeitschr., 1894-95, 108. 1 8 ELECTROLYSIS OF WATER. OUTPUT. The theoretically possible output would be 626.4 X 300 X 44X36 = 297.6 cubic metres of gas. Latchinoff has, there- fore, assumed a theoretical current output and has reckoned out with reference to the chosen voltage used an energy out- put of 57.6 per cent. COST OF OPERATION. Ivatchinoff has also given estimations of the working cost of his process. In order to be able to compare the calculated cost of production, we will assume always two extreme cases, namely, when power costs on the one hand % cent per kilo- watt-hour for water power, and on the other hand i J^ cents for steam power in small plants. The data of Latchinoff would then reckon up for the manufacture of 100 cubic metres of oxygen and 200 cubic metres of hydrogen in thirty- six hours as follows : 1 188 kilowatt hours $2.97 $14.85 Caustic soda 0.08 0.08 Drying of gases 0.24 0.24 Attendance 4.05 4.05 Total 7.34 19.22 Therefore, according as to whether only the oxygen, or only the hydrogen, or both gases are assumed as salable, we make these calculations : a. With water power, i cubic metre of oxygen costs 7.34 cents. i cubic metre of hydrogen costs 3.67 cents, i cubic metre of detonating gas costs 2.45 cents. b. With steam-power, i cubic metre of oxygen costs 19.22 cents. i cubic metre of hydrogen costs 9.61 cents, i cubic metre of detonating gas costs 6. 40 cents. Sinking fund and interest on the capital invested are not included in these figures. PRACTICE. It is not known whether plants after Latchinoff's methods have been industrially operated. The apparatus is said to have been publicly exhibited in the IV Blectrotechnical PROCESS OF DUCRETET. Exposition, opened the 24th of January, 1892,' at St. Pete s- burg. The baths were, however, injured by an accident at the exposition, so that instead of being in continual operation, as was intended, they were only operated a short time. Several French investigators were working on the problem of the technical electrolysis of water simultaneously with Latchinoff, and, according to all appearances, independent of him. Process of Ducretet, 1888. Ducretet, for instance, constructed many forms of apparatus for the decomposition of water, using alkaline electrolytes. FORM OF APPARATUS. One of these forms of apparatus is shown in Fig. 8. The iron cyl- inder E' serves at the same time as an electrode and container for the electrolyte. The bottom i is covered with an insulating layer and bears the supply tube T'. The cover Co is insulated and is screwed on the tube. One of the connections, , is united with a centrally placed cylin- der made of iron wire netting, which forms the second electrode. The cylindrical asbestos diaphragm S S divides the tube into an anode and a cathode space, and the gases es- cape at T' and G'. 2 Process of Renard, 1888 to 189O. The construction of Renard rests on a similar principle. He began his work in electrolytic decom- position in 1888 and communicated his results obtained in a stance of the 1 Chem. Ztg., 1892, 28, 461. Elektrochemische Zeitschr, 1894-95, 106. 2 La lumire lectrique, 4O, 234. 20 ELECTROLYSIS OF WATER. French Physical Society, December 5, 1890. Renard took up the question primarily as commander of the balloon corps at Chalais, and, in consequence of this, had in view a simple and cheap method of obtaining hydrogen for the filling of balloons* OUTPUT. Renard gives the following output as obtained in a practi- cal application of his apparatus. 1 Volume of hydrogen at ioC and 760 mm. 0.433 litre, per ampere-hour. " " 0.144 litre, per watt-hour. Ampere-hours consumed per cubic metre of hydrogen, 2310. Watt-hours " " " " " 6930. H. P. hours " " " " " 9.4 Renard assumed, therefore, nearly a theoretical current out- put, and at a working voltage of 3 volts, a useful energy effect of about 50 per cent. He worked on the theory that the platinum used in the laboratory apparatus for the decomposition of water may be replaced by a cheaper material, and, besides that, he tried to find a diaphragm which would separate the gases without in- troducing too much resistance. In consequence of these endeavors, he replaced acid electro- lytes by alkaline solutions and hereby was enabled to use cast- iron or steel as electrode materials. Since clay diaphragms showed too high a resistance for the purpose in view, Renard used asbestos as a diaphragm material, its resistance being practically negligible. As an electrolyte, he used a 13 per cent, solution of caustic soda, the resistance of which is the same as that of a 27 per cent, solution of sulphuric acid. FORM OF APPARATUS. The simplest construction used at Chalais consisted of a large cylindrical sheet-iron vessel, which acted simultaneously as cathode and a container for the electrolyte. A perforated sheet-iron cylinder which hung from the insulated cover of 1 La lumiere electrique, 39, 39. PROCESS OF DUCRETET. 21 the vessel, served as anode. An asbestos sack drawn over this served as a diaphragm. This consists of exactly the same principle of construction as was seen in the apparatus of Ducretet. Renard showed to the French Physical Society two types of apparatus, one weighing 2 kg., using a current of 25 am- peres at 2.7 volts, while the second was intended for a current of 365 amperes at 2.7 volts, and furnished 158 litres of hy- drogen per hour. The price of an apparatus of this latter size was given as $20. PRACTICE. Such an apparatus was in operation at Chalais for six months, and at the end of this time both electrodes and dia- phragms were in the best of condition. According to the report of Renard the hydrogen obtained was pure and the oxygen free from ozone in consequence of the alkaline solution, so that the use of rubber for connec- tions was permissible. The gases were washed with tartaric acid in order to remove the spray of the alkaline solutions. COST OF PLANT AND OPERATION. The cost of a plant which used 36 of the larger-sized elec- trolyzers, and furnished 5.7 cubic metres of hydrogen and 2.85 cubic metres of oxygen per hour (=136 cubic metres of hydrogen and 68 cubic metres of oxygen per 24 hours), was estimated by Renard to be $8100 and the cost of operation, including the compression of the gas to 120 atmospheres, 10 to 12 cents per cubic metre of hydrogen. It was assumed in these figures that 2 x / 5 pounds of coal would be used per H. P. hour in generating current. LABORATORY APPARATUS. Besides the apparatus for the technical manufacture of hy- drogen for the filling of balloons, Renard constructed also larger laboratory apparatus which were then furnished by Du- cretet in Paris 1 . 1 La lumiere lectrique, 43, 432. 22 ELECTROLYSIS OF WATER. These apparatus were very similar to the above-described Ducretet apparatus and are shown in Fig. 9. Fig. 9. This laboratory model measured 40 cm. high, 18 cm. diam- eter, and could take a current of 60 amperes using 4.5 volts. The production was 26 litres of hydrogen and 13 litres of oxy- gen per hour. With a normal tension of three volts the ap- paratus passed 25 amperes, producing n litres of hydrogen and 5.5 litres of oxygen per hour. The materials used and the electrodes were the same as in the formerly described ap- paratus of Ducretet. Instead of asbestos diaphragms clay cylin- ders were used, in order to obtain very pure gases. In this case the cylinders are perforated at the bottom and contained a bent tube, S, which connected the two electrode chambers. The electrolyte was poured in at M. The two gas bottles PROCESS OF DELMARD. Gh and Go served to equalize the differences of pressure, an contained a 5 per cent solution of tartaric acid to absorb the alkaline spray carried over. As the current density increased, the voltage used and the output increased in the following manner : TABLE II. Amperes. Volts. Litres hydrogen per hour. Litres oxygen per hour. Temperature. 2 2.06 0.87 0-43 25-5 5 2.24 2.16 1. 08 10 2.41 4-33 2.16 20 2.84 8.66 4-33 25 3-4 10.82 5-41 40 3.65 17.32 8.66 5 4.00 21.65 10.82 60 4.40 26.00 13.00 Process of Delmard, 189O. The principle of construction as used by Ducretet and Ren- ard was patented in Germany by Delmard in the German pat- ent 58,282 of November 23, 1890. The patent claim embraces : 1. "An electrolytic apparatus for the decomposition of wa- ter consisting of an iron vessel, g^ in which an iron tube, #, provided with perforations, 0, and surrounded by a sack of as- bestos cloth, is hung from the cover /, and insulated there- from, said cover being provided with two pipe connections for the taking off of the hydrogen and oxygen, and so arranged that the outer vessel forms the + and the inner tube h the - electrode." (See Fig. 10.) 2. " In combination with the apparatus described under i, a pressure regulator consisting of the vessels A and B con- nected by a tube, T, close to their bases, into which vessels the hydrogen and oxygen coming from the generator are led by the tubes O and H, which dip to the same level in the two vessels, atmospheric air being excluded." (See Fig. n.) It is not necessary to further describe these patent specifi- cations since they furnish nothing new compared to Ducretet and Renard. The apparatus is merely as compared with these latter considerably longer. ELECTROLYSIS OF WATER. Fig. 10. It is quite probable that this patent is only a German patenta pplication of Ren- ard's under a strange name. Nothing is known of any industrial ap- plication of this type of apparatus, espe- cially in Germany. Process of Bell, 1893. We first find something new in appa- ratus for the decomposition of water, using Frg. ii. porous diaphragms in the German patent 78,146 of October 30, 1893, describing the apparatus of Bell. 1 PATENT CLAIM. Patent claim : u An arrangement for the continuous charging of apparatus for the electrolytic decomposition of fluids, in which the elec- trodes are separated from each other by a partition, consisting therein, that these partitions themselves, or an extension of 1 Extracts in Zeitschr. Elektrochein., 1894-95,429. Jahrbuch f . Elektro- chemie, 1895, 193. Ahrens: Handbuch der Elektrochemie, 1896. PROCESS OF BEIvL. 25 the same made of capillary netting or a similar material, have their upper edge turned over and dipping into the vessel g g above the decomposition space and filled with an electrolyte or another suitable fluid by a funnel, z, and out of which the electrolyte or other fluid is sucked up, and by the proper reg- ulation of the level of the fluid in the vessel g g furnishes to the electrodes, by the principle of capillarity and gravity, a fresh quantity of fluid as the latter is decomposed." FORM OF APPARATUS. The apparatus, as shown in Fig. 12, contains on a base 1 plate,/, two concentric hollow cylinders, ^and c^ of cast-iron, which are described in the patent specifications as being of 26 ELECTROLYSIS OF WATER. circular section. Both cylinders are insulated from each other by a non-conducting clamp, f, and covered by an insulated cover, d. The funnel g passes through the middle of the cover which contains an inverted conical tube, r. An asbestos web, s, in which is interwoven vertical vegetable fibres of linen or wool, hangs between the electrodes. The lower end of the cylindrical diaphragm is fastened to the hard rubber clamp surrounding the one electrode holder, while the upper end is turned over the tube r. The funnel-shaped vessel Discovered with an air-tight cover, in the center of which a funnel-shaped tube, t, is fastened by means of a stopper, and reaches down al- most to the upper end of the electrode C. The outer contain- ing vessel C carries the nipple D, on the end of which, as well as on the end of the tube ^, the rubber tubes serving to carry off the gases may be slipped. A copper band, , which sur- rounds the outside hollow cylinder, takes the current to the negative pole, while the positive conductor is connected to the inner cylindrical electrode by means of the plate a, in contact with the base plate/. The spaces between the electrodes and the diaphragm were filled to n with pieces of cast-iron or some other broken conducting material which is not attacked by the electrolyte. This makes a coherent conducting layer which reaches to the diaphragm. This direct contact of the electrodes with the diaphragm is obtained by Bell by still other arrangements, such as by two narrow spirals of steel wire upon which the asbestos mantel is clamped. In this case the inner cylinder can be completely omitted while the outer cylinder serves as a container. In this the fundamental condition is always that the gases evolved must be able to escape easily. The electrolyte is poured in through the siphon-shaped fun- nel i connected with g. A 15 per cent, solution of caustic soda is used for electrolyte. The electrolyte is run in until it flows over the edge of r and runs down the asbestos mantel into the apparatus. After that the asbestos mantel absorbs PROCESS OF SCHMIDT. 27 the electrolyte constantly by capillary action. The height of the electrolyte can be observed in the tube k T . Fig. 13 shows a cross-section through the two electrodes and the diaphragm, while Fig. 14 shows another arrangement Fig. 13. Fig. 14. which gives to the diaphragm and therefore also to the elec- trode surfaces a greater area. Although the thought of obtaining the greatest possible electrode surface by this filling with metallic granules is quite reasonable, yet we must not leave out of consideration that the electrodes may be coated over by impurities which sepa- rate out during electrolysis, and thereby produce difficulties in the operation. PRACTICE. It is not known that this apparatus has gone into prac- tical use. The whole construction reminds one more of a simple laboratory research, as seems particularly indicated by the three set-screws F which are only used upon instruments requiring fine adjustment. These are for the purpose of making an exact horizontal adjustment of the edge of the diaphragm above r. After the above-described attempts of Bell, practical electro- chemists, who worked on the electrolytic decomposition of water, turned to other principles of construction which will be described in the latter part of this work, while the use of porous diaphragms fell more into the background. Process of Schmidt, 1899. It was not until the year 1889 that Dr. O. Schmidt returned to the use of porous diaphragms and obtained a patent for the 28 ELECTROLYSIS OF WATER. apparatus described in the German patent 111,131, June 13, 1899.' PATENT CLAIM. Patent claim : "An apparatus for the electrolysis of water consisting of many series of plates after the fashion of a filter press, and characterized by having the pipes for conducting off the gases communicating with the water supply pipes, for the purpose of taking back into the electrode compartments the water car- ried out by the gases." DESCRIPTION. The fundamental idea of the Schmidt invention consists, therefore, in an apparatus with numerous cells placed like a filter-press and characterized by the fact that the current com- municated to the water in the direction of the gas exits, by the evolution of the gas, causing a large outflow of water, is converted into a rapidly circulating system. This is obtained by making the gas exit pipes connect with a vessel called the gas separator which connects with the pipes furnishing the cells with water. Since the water carried out by the gases is thus returned to the electrolyte in the cells, very little loss of electrolyte ensues. The apparatus is shown in one example in Figs. 15-18, in the form in which the electrodes are connected in series. Fig. 15 shows the apparatus in side elevation, Fig. 16 in horizontal section, Fig. 17 a front view of one of the frames as seen from the anode side, and Fig. 18 is a vertical middle section through the gas separator O for the oxygen. , e are the double pole electrodes touching each other at the thick edges, and d the diaphragms placed in between the electrodes to produce the electrode spaces and simultaneously insulate the edges of the electrodes from each other. Each plate, 5~ 6.75 i cubic metre of oxygen 2.70-13.50 when using iron electrodes. Using acid electrolyte (sulphuric acid of a specific gravity of 1.235) and lead electrodes, Schoop 1 says, from the data ob- tained in operating a plant in question, that a voltage of 3.9 is needed with cold acid and 3.6 with warm acid. From this 1 Private communication to the author. PROCESS OF SCHOOP. 57 data, assuming a theoretical current output and 3.5 volts to be used, there is required 18.67 kilowatt hours for producing 3 cubic metres of mixed gas, and the cost of production with power at ^ and i j{ cents per kilowatt hour is respectively : i cubic metre of detonating gas I -55 < >- 7- 750 cents. i cubic metre of hydrogen 2.325-14.625 " i cubic metre of oxygen 4.650-23.250 " PRACTICE, The Schoop water decomposition apparatus was tried in an experimental plant of small capacity by the Cologne Accumu- lator Works of G. Hagen, at Kalk on the Rhine, and this firm erected a plant for a total daily production of 15 to 20 cubic metres of hydrogen and 7.5 to 10 cubic metres of oxygen. Fig- 33- The plant consisted of 18 cells connected in series, through which was sent 150 to 200 amperes. The switchboard con- tained: i Weston ammeter for 250 amperes, i voltameter, i indicator of the direction of the current, i low current cut out, i signal bell, and 2 water manometers. The current was fur- 58 ELECTROLYSIS OF WATER. nished by a shunt-wound dynamo of 250 amperes and 65 volts. The primary cost of the plant was approximately $1,250, and the operation of the same resulted in a saving of $1,750 yearly, as compared with the previous method of manufacture of hy- drogen for soldering purposes. Fig. 33 shows part of this plant. The same firm has undertaken the construction of the Schoop apparatus for foreign customers. Process of Hazard-Flamand, 1898. If we do not confine ourselves exclusively to the electrolysis of water, the Hazard-Flamand apparatus may be classified with this group of apparatus. It was patented in Germany, No. 106,499, of June 12, iSgS. 1 FORM OF APPARATUS. The inventor describes his construction as a " water-sealed diaphragm " for electrolytic apparatus, especially for the elec- trolysis of water. The principal idea of the diaphragm, which is intended for the general manufacture of gas, and which pre- vents mixing of the anode and cathode fluids saturated with their respective gases, is shown in Figs. 34 and 35. In the box A, which forms the cathode and is provided with ribs, #, for increasing the lateral surface are transverse parti- tions, #', cast in one piece with the box A, and likewise ribbed a z for increasing the surface. The deep gutter B is filled with an insulating fluid and receives the edge of the cover M. The cover is insulated from the vessel containing the electrolyte by the ebonite blocks B'. The clamping screw D conducts the current. In the chamber formed by the transverse parti- tions a' is the anode E, likewise provided with ribs e ; e 1 is the conductor which is led through the cover and insulated therefrom. The anode is arranged in a bell, 2 , and projects into the cylindrical holder m z . The latter is in one piece with the cover and closed in, gas-tight, with paraffin. The several pos- itive electrodes are naturally connected in parallel. The an- i Zeitschrift f. Elektroch., 1899-1900, 511. PROCESS OF HAZARD-FLAMAND. 59" odes are surrounded by diaphragms, which consist of long gut- ters, H, of Y-shaped section made of ebonite, porcelain, glass, etc. The upper parts of these run underneath, almost touching the bottom of the next gutter below. The rings- Fig. 34- have attachments, h (Fig. 34), in order to hold them tightly in place. The lower ring rests with its side projections upon the insulated cap /, which latter has the function of prevent- ing hydrogen evolved from the bottom of the vessel from pass- ing into the anode chambers. The uppermost ring //is of a different shape from the others for the purpose of closing per- fectly the anode cells. The cover M projects downwards into the uppermost gutter h' by means of the circular flange m~ 6o ELECTROLYSIS OF WATER. Fig 35 The gases are collected in two canals. A", in the cover, con- nected by suitable openings, ;/, in the respective cells. The pipe connections branch from these canals. The flanges m 1 in the cover are the containers which communicate with the inside of the decomposition vessel and are rilled with elec- trolyte. This prevents escape of gas from the openings h* and m 3 . While the apparatus is in action, if the level of the fluid in m 1 varies through inattention, the gas will escape into the air through ; 3 , but without mixing in the apparatus. Like- wise, no mixing of gas takes place if the internal pressure rises because of irregularities in the conducting off of the gas,, as in that case the gas escapes from the apparatus as soon as the level of the liquid has sunk to the lower edge of the bell m. PROCESS OF GARUTI. 6 1 PRACTICE. Nothing is known of any industrial application of this some- what complicated apparatus. Process of Verney, 1899. A similar idea is embodied in P. J. F. Verney's invention, French patent 280,374*. This apparatus is intended primarily for the electrolytic de- composition of water, but is also suitable for the decomposi- tion of other electrolytes, for the purpose of obtaining gas. FORM OF APPARATUS. The apparatus consists of a metallic box, divided above into a number of chambers by a series of partitions cast in one piece with the external shell. The external vessel with its partitions serves for one pole and the other pole is formed of metallic sheets contained in the compartments of the box. In- sulating rings of Y-shaped section are arranged around these grooved sheet electrodes, the downward-plunging part of each ring extending into the cover-shaped enlargement of the ring beneath, without, however, touching it. By this means suf- ficient permeability section for the current is provided, with at the same time complete separation of the gases. This is therefore a patenting of the same subject under two different names. PRACTICE. Verney's apparatus had no application in practice at least as far as the author can learn. Both of the latter described apparatus remind one in many ways of the curtain diaphragms such as have been proposed for the technical electrolysis of the alkaline chlorides. + a -L * M N Fig. 36. Fig. 37. Fig. 38. tween a and b is correspondingly reduced, this evolution of gas occurs only at the end electrodes and not upon the metallic partition. The partition will, therefore, act only as a bi- polar electrode, when we employ more than 3 volts tension between the poles a and b. The metallic partition, therefore, takes no part in the electrolysis as long as the tension used is not very much greater than that needed for a single decom- position cell. According to a recent communication of Buffa 1 , the priority of this observation is due to the Italian Del Proposto, while Garuti, in company with Pompili, applied the idea technically for the construction of various forms of apparatus for the elec- trolytic decomposition of water. According to the patent application of Garuti, previously alluded to, the fundamental form of the apparatus may be de- scribed in the following words : 1 Bulletin de 1'Associat. des Ing. Electr., 11, 305 (1900). 6 4 ELECTROLYSIS OF WATER. FORM OF APPARATUS. u A practical form of apparatus characterized by a holder or container in the form of a gasometer bell dipping into the electrolyte and provided with two collecting chambers sur- rounding the electrodes, and to the walls of which are soldered metallic diaphragms parallel to each other and hermetically tight, for the purpose of accelerating the separation of the Figs. 39 and 40. evolved gases from the electrodes, to give to the gas the*nec- essary pressure for their use and to make irregularities in their production at once perceptible." Garuti used, in his first apparatus, electrodes, supports and even vessels of lead. The drawings of the first Garuti apparatus referred, as far PROCESS OF GARUTI. 65 as can be learned from the patent description, to a model left with the German office, suitable for a current of 17 amperes. The apparatus as shown in Figs. 39-44 consists of a wooden box A lined with sheet lead a. This box holds the electrolyte Fig. 41. Fig. 42 Fig. 43- as well as the electrolyzing apparatus itself, and stands upon insulators M. Twelve per cent, sulphuric acid is used as elec- trolyte with lead electrodes. The real electrolyzing apparatus consists of a rectangular box open beneath, Fig. 43, divided by metallic plates into long, narrow divisions n. These plates form the metallic walls for the separation of the gases. The electrodes are introduced into the chamber thus formed by means of the wooden comb Fig. 44, which serves at the same time for insulating the 66 ELECTROLYSIS OF WATER. electrodes from each other and the metallic partitions. Fig. 40 shows best in its horizontal section the arrangement of the electrodes, their connection with the main electrical conduc- tors, as well as the position of the partition walls, n being the metallic diaphragms, cC the one pole, b the end of the other pole. The distance of the electrodes apart is 10-20 mm., the height of the electrodes 140 mm.; the anodes are somewhat heavier than the cathodes to allow for the production of super- oxide, the former being about 3 mm. in thickness, and the latter i mm. The electrodes are prevented from coming in contact with the bottom of the inverted diaphragm chest by means of fork- shaped pieces of wood resting upon (Fig. 44) similar perfo- rated pieces of wood. Naturally the electrodes stand free from the end surfaces of the box. Likewise also, the connecting pieces O (Fig. 42), which con- nect in parallel the electrodes of like polarity, must be insula- ted from the back walls of the diaphragm chest. When the in- sulating parts are removed, both systems of electrodes may be easily lifted sideways out of the diaphragm chest. The conductors are fastened to the screw- clamps b 1 b 2 (Fig. 41). The spaces formed by the diaphragms are alternately perforated and their exits are united with two parallel-arranged gas collectors G (Figs. 41 and 43). The height of these chambers, which serve likewise to regulate the pressure, must be changed with the pressure desired. The gases pass from these chambers by the lead tube 6*. Above these exit tubes are wider tubes ^, made of insulating material (porcelain or glass), and the main gas conductors connect only to these in order to insulate the apparatus electrically. A Fig. 44- PROCESS OF GARUTI. 6 7 lead tube H, likewise soldered to the gas chamber allows[a gas- tight joint to be made by filling with water between H andlS. The whole system of electrodes and partitions rests upon two wooden stringers 6 cm. high in order to avoid any short circuits in consequence of the falling down of peroxide and stopping up of the apparatus by precipitation from the elec- trolyte. Two iron handles, L (Fig. 41), covered with lead, permit the whole system to be raised out of the box. Two amperes per sq. dm. is described as a suitable current density for dilute sulphuric acid of the above given concentration. VARIATION OF APPARATUS. Garuti describes in his German patent still another modi- 45 Fig. 46. fied form of his apparatus which makes a separation of the gases good enough for industrial purposes. This is shown in Figs. 45 and 46. The most important difference between this modification and the previously described apparatus consists in an enlarge- 68 ELECTROLYSIS OF WATER. ment of the distance between the electrodes to about 30 mm. and an alteration in the form of the partitions. The latter are not at the same height in the whole apparatus, but diminish towards the middle of the apparatus so that the effective elec- trode surface is increased. This results naturally in a greater danger of the formation of mixed gas, which, however, is taken care of by the greater distance between the electrodes and the higher water pressure. The chamber G must there- fore be correspondingly deeper. By this arrangement Garuti intends to use a current density of 4 amperes per sq. dm. with a voltage not higher than in the type of apparatus previously described. The testing of the Garuti cell in practice, and the experi- ence accumulated therewith, made the introduction of various improvements necessary. According to communications in the technical journals of 1899*, Garuti has dropped the use of lead for the electrodes and partitions. These changes are rendered necessary by the irregular action of the formation of superoxides and particularly also by the weakness of the lead which, by getting out of shape, made very disagreeable short circuits possible in such an apparatus, with very narrow chambers. Garuti therefore passes, as other inventors before him, to the use of iron. This requires also a change in the electrolyte, which Garuti likewise changed to caustic alkali solution. The improved apparatus (Fig. 47-50), retains its general form, but the outer box as well as the electrode system is of iron. The distance of the electrodes from each other is decreased to 12 mm. in order to increase the efficiency of the apparatus while the distance of the lower edge of the electrodes from the bottom of the containing vessel is reduced to 12 cm. The partitions are no longer entire but have a zone of fine perfo- rations 4 cm. wide placed by Garuti at first at the lower edge 1 L'Elettricita, 1899, 37, 502 ; L'Ind. lectro-chim. 11, 113 (1899). PROCESS OF GARUTI. 6 9 of the electrodes, according to Fig. 48, and later at the level of the middle of the electrodes, as shown in Fig. 49. These two alterations form the substance of the German h nA OOOOOODOO O O OO O G CO O O O O O O-5 o o o o o oo Fig. 48. S Fig. 47- patents 83,079 and 106,226, which, however, are not taken out in the name of Garuti, but by the German license the So- cie'te anonyme 1'Oxyhydrique, of Brussels. The correspond- Fig. 49- Fig. 50. ing English patents are those of Garuti and Pompili, num- bers 23,663/96 and 12,950/1900. In the latter patents Garuti also describes a new form of diaphragm chamber, the length of which in the older type of apparatus is often two metres, causing considerable difficulty in its manufacture. ELECTROLYSIS OF WATER. The new manner of manufacture and connecting up the diaphragm chamber consists in the following : A rectangular sheet of the length of the electrolyzer and about twice as high is cut through along the middle line b c (Fig. 51) somewhat over half way, and thereby divided into three parts efg. The two laps ^/are now bent over along the lines/ i h at right angles, but in opposite directions. The sheet represents thereby the shape shown in plan in Fig. 52, in section in Fig. 54, in perspective in Fig. 53. / c __ _\ __Z c / a> a r Fig. Fig. 52- This forms two cell-like spaces, m and /z, each having a length of half the sheet used. In building up the diaphragm chest from such pieces, each edge k of one piece is connected with the edge / of the next box. The combination of two such elements x y, is shown in Figs. 55 and 56. The line 2 is u 1 2 R 1 Fig. 53- Fig. 54- the point of union of the two cell elements. The diaphragm chest (Fig. 57) obtained in this way contains, therefore, cells, which, considered from the middle line r j, are open above on one side and closed above on the other. Then all the cham- bers m contain, for instance, anodes and all the chambers n cathodes ; thus the gas escapes separated at both sides of the apparatus. By this arrangement of the diaphragm chest, Garuti reaches a simplification of the points of contact com- pared to his older construction. PROCESS OF GARUTI. All insulating parts are of asbestos, which is not attacked by the dilute caustic soda solution. Lead and zinc soldering is to be avoided. The partitions project below the electrodes at their under \ / I ! k / \ / I ' ! ' i \ I ^ / \ I / I \ V* xg v -\ >A ; I i / \ / "~~~V \ N ! 1 / / \ 7 vl A \e obtained in the former, with a current strength of y 2 am- pere, 93.9 per cent, of the gas obtained in the latter; with i ampere 89.4 per cent., with 2 amperes 87 per cent., 3 am- peres at first 87 per cent, then 90.8 per cent., and still later 93.6 per cent. The rise of the percentage with the last cur- rent strength is explained by the rise of the temperature of the voltameter to a condition of equilibrium in the last case and at high temperatures the formation of persulphuric acid diminishes. Using the acid of specific gravity of 1.15, the acid voltameter gave very accurate results. The test was made by a current of 3 amperes corresponding to a current density of 4 amperes per sq. cm. The above averages are shown in the following table : TABI,E vi. Current strength. Tension. Current den- sity per sq. cm. Remarks. 0.2 amperes 0.5 0.92 2.42 " 4.21 1.85 volts I. 9 8 2.0 9 " 2.32 2-55 " 0.25 amperes 0.62 " 1.15 " 3.02 " 5-26 The ordinary voltameter filled with sulphuric acid (with an anode sur- face of 7 sq. cm.) required on the other hand 3 to 3. 5 volts tension. VOLTAMETER OF WAITER-NEUMANN. The voltameter of Walter-Neumann (Fig. 75) consists of a glass bulb, to which is attached a graduated glass tube with a glass stop-cock and funnel. The bulb contains the two plati- num electrodes, which are connected with the source of cur- rent by platinum wires sealed through its sides. A rubber connecting tube communicates with the straight, ungraduated, glass tube carried by the same stand. In using the instrument the filling tube is raised with the stop-cock of the graduated tube left open until the acidulated water fills the latter, upon which the stop-cock is closed and the current turned on. Working in this way, the irregular interval at the beginning of electrolysis falls within the time 9 6 ELECTROLYSIS OF WATER. . 75- Fig. 76. of the experiment. More accurate results are obtained if the filling tube is left with stop-cock open until the electrolyte in the graduated tube stands at the upper point, and the cur- rent then be turned on, the cock being still open, and the lat- ter only closed when the current has reached a constant value,, and the time of the experiment reckoned from this instant.. During the electrolysis the filling tube is constantly lowered so that the level of the liquid in the graduated tube and the filling tubes are at the same height at the end of the experi- ment for the purpose of making the reading. The two glass tubes should be of nearly the same section in order to equalize the effect of capillarity, which is scarcely noticeable except in. too narrow tubes. Fig. 76 shows a voltameter on the same principle. It is distinguished by having the funnel closed by a ground stopper PROCESSES AND APPARATUS. 97 carrying a thermometer, and the leveling of the fluid in the two tubes is made, not by lowering the filling tube but by drawing off the electrolyte. The support carries an adjust- able mark for indicating the level. It is not possible with this apparatus to equalize irregularities at the beginning of the electrolysis. VOLTAMETER OF BERTIN. The voltameter of Bertin employs the hydrogen alone for the measurement. It consists of a graduated burette, con- nected at its upper end by a capillary tiibe with a glass bulb. The latter carries a hose connection in order to be able to draw electrolyte into the burette. The burette dips into a vessel which con- tains the two platinum elec- trodes, of which only the neg- ative one reaches into the bu- rette. The capillary tube hinders the rise of gas bub- bles into the gas bulb, so that the evolved hydrogen can be read off directly. The equalizing of the level of the liquids is done by the proper immersion of the burette. As far as concerns the irreg- ularities at the beginning of the experiment, the same suggestions apply to this as to the previous arrange- ments. Fig- 77- 9 8 ELECTROLYSIS OF WATER. VOLTAMETER OF MINET. Two voltameter constructions of Minet 1 are different from those already described. One arrangement, which is shown in Figs. 77-80, is intended to avoid the sources of error in the mixed gas voltameter, such as absorption of gas and secondary reactions. The apparatus consists of two principal parts. In the up- . 78. Fig. 79. Fig. per right-hand part are two decomposing vessels, C t C 2 , which communicate with each other by the three-way stop-cock R 2 . To the left are two communicating tubes. The tube t is graduated and is in combination with the decomposing space C x , by means of the narrow, bent tube O O t . The electrolyte is poured in through the rilling tube connected with the de- composing vessel C 2 , in charging which the stop-cock R 2 has the position shown in Fig. 78 and the stop-cock R i? is left open. The electrolyte is run in to the marks O x O a . Then the tubes IF are filled to the marks O O 1 , during which the the stop-cock R t is left open and R 3 is closed. The latter has the object of exact adjustment of the level in the communi- cating tubes. At the beginning of the electrolysis the electrolyte must be exactly adjusted to the marks O 1 O O x O a . Fig. 80 shows the position of the three-way cock R 2 as soon as the electrolyte is to be replaced for another measurement. 1 Trait^ thoriqueet pratique d'Electro-Chimie, 1900, 353. PROCESSES AND APPARATUS. 99 The decomposing vessel C 2 contains only one electrode A a , while the decomposing space C, contains two of them A and Aj. This arrangement allows either mixed gas or oxygen or hydrogen to be caught as desired. The determination is only made when the current has become constant. This is quickly reached when the source of current has a sufficiently high electromotive force, and its strength is correspondingly regu- lated by a rheostat. During the beginning of the electrolysis the stop-cock R t remains open. Closing the same is the beginning of the ex- periment, opening the switch is the end . The equalizing of the pressure in if, necessary for reading the gas volume, is done by the manipulation of the dropping stopper R 3 . The water used is saturated with oxygen and hydrogen before the ex- periment and is acidulated with 0.5 per cent, sulphuric acid. According to Minet's researches, which are in part repro- duced in the following table, the method of working described is suitable only for currents under 0.5 ampere. TABLE VII. Calibration of a galvanometer. Inner resistance at 18 C. = 17.1 Ohms. Inserted Currents trength Current strength in the Number of Galvanometer constant. resistance. in the voltameter. J galvanometer. divisions. n * = ! n Ohm. Ampere. Ampere. Ampere. 00 0.00369 0.00369 9- 6 5 0.000382 < < 0.00859 0.00859 22.50 0.000382 ' ' 0.01464 0.01464 38-50 0.000380 4.00 0.07608 0.01415 37.00 0.000382 " 0.0997 0.01897 49.80 0.000381 o. 1098 0.02097 55-00 0.000381 -734 0.1197 0.00496 13.00 0.000381 ' ' 0.1206 0.00498 13.00 0.000383 2.OO 0.1206 0.01272 33-00 0.000385 0-734 0.1585 0.00656 17.00 0.000386 11 0.^79 0.01155 30.00 0.000385 ' ' 0.417 0.01722 45.00 0.000382 i < 0.466 0.01925 50.00 0.000385 < < 0.557 0.02296 60.00 0.000383 Minet claims for his apparatus the following : (i) The decomposing spaces are independent of the rest of the apparatus. Their volume can be reduced at pleasure. The electrolyte is therefore quickly saturated with gas. 100 ELECTROLYSIS OF WATER. (2). If the pressure is kept constant during the experiment by manipulating the stop-cock R S there need be no fear that the volume read off will be too large by reason of the setting free of absorbed gas. (3). The volume of the tube between O O x , is so small that no correction is necessary for differences of temperature. . (4). The graduated tube / consists of several parts: Of two en- largements (P P x ) of different content, and a straight tube which is divided in o.oi ccm. The divisions are 2 mm. apart. The enlargements allow the measurement with the one apparatus of widely varying current strengths without giving to the graduated tube abnormal dimensions. (5). Since t and t f are of like calibre, no correction is neces- sary because of capillarity. Using electrolytes with 0.5 per cent, sulphuric acid and a cur- rent strength below 0.5 ampere, there is no loss by the for- mation of ozone, hydrogen dioxide, or persulphuric acid, as well as no loss of hydrogen noticeable by reduction of these products. These accessory phenomena only begin with a higher acid concentration and a greater current strength. VOLTAMETER OF MINET FOR INDUSTRIAL PURPOSES. Minet's voltameter for industrial purposes 1 rests on another principle of construction. In this arrangement the increase of pressure upon an en- closed volume of gas is measured. A glass vessel with a contraction O, as in Fig. 81, stands in combination with a manometer M, which is constructed to stand a pressure of two atmospheres and allows reading to i/ioo of an atmosphere. A is a tube which may be her- metically closed and through which the electrolyte may be poured in. The lower compartment of the glass vessel contains the electrolyte, the level of which is raised at the commencement of each measurement to the mark O so that the upper part of the apparatus always contains the same gas volume V. 1 A. Minet: Traite theorique et pratique d'Electrochemie, 1900, 357. PROCESSES AND APPARATUS. 101 Fig. 81. The increase of pressure shown by the manometer is / = K t z, (i), in which v is the volume of the gas evolved during the meas- urement reduced to normal atmospheric pressure P; K x is a coefficient depending upon the volume V and the pressure P. Therefore we have and on the assumption that the temperature during the experi- ment remains constant. Since, on the other hand, the volume V of the gas evolved is proportional to the current strength Q, therefore 102 ELECTROLYSIS OF WATER. and putting for v its value from the first equation, we have P Q = KK and if we place ^ ^ = K, we have Q = K./ ; that is, the cur- 12 rent strength going through the galvanometer is proportional to the pressure shown by the manometer. The coefficients K x K 2 which limit the value K can be cal- culated if we know the pressure at the start, and the tempera- ture, which is taken as constant for the duration of the ex- periment. Usually the coefficient K is determined experimentally by a comparison of the manometer results with those of a cor- rectly calibrated galvanometer, and correctly measuring the time of the electrolysis. The voltameter naturally gives accurate readings only when the temperature and pressure are the same as when the instru- ment was calibrated. But since the temperature in laboratories varies very little and the atmospheric pressure seldom more than i per cent, above and below the normal, the instrument is sufficiently exact for a series of practical tests with small current strength ; for example, for rough electro-analytical work where it can be used likewise as a current meter. For a volume of V 1500 ccm. and an allowable increase of pressure of 3 atmospheres, the instrument will register 4-5 ampere hours. () For Technical Purposes. Process of Eldridge, Clark and Blum, 1898. For the technical manufacture of mixed gas (oxy-hydrogen gas) we know of only one, that of Eldridge, Clark and Blum, U. S. patent, 603,058. We have not heard of the practical application of this pro- posal, and since such is hardly to be expected, it will suffice if we here, 'for the purpose of completeness, reproduce extracts PROCESSES AND APPARATUS. 103 from the patents which have appeared in the technical journals. 1 The whole arrangement is shown in Figs. 82-84. A steel cylinder, 2, provided above and below with flanges, 5 and 6, is lined inside as thickly as possible with a coating, 70, of graphite or some difficultly fusible material. The bot- tom of the space inside is formed of the carbon cathode 7j, Fig. 82. fastened tightly to the steel floor-plate j. The latter arrange- ment is held tightly down upon two stoneware plates, 8 and p, by bolts 7. There are diametrically opposite openings, 77 and 12, through the two walls of the cylinder, the first for leading in water and the last for discharging the gas formed. The inside of the cylinder is closed air-tight by a || shaped cover made tight by an asbestos gasket, / Co. and Obach. j 750 3-0 2.25 187.50 83-25 El. A. G., formerly ) Schuckert and Co. ] 600 2-9 1.74 62.50 36.00 must not be overlooked that with bipolar connections a con- siderable amount will be saved in piping, electrical conductors, and erection. This difference is most easily seen when we consider, for instance, the operation of the plant for the filling of air balloons. A normal military balloon requires, in order to be filled in twenty-four hours, a plant of about 200 kilowatts. To do this, 6 or 7 units of the Schmidt apparatus of the largest type will be necessary, while 320 of the Schoop electrolyzers, 200 of the Garuti, 90 of the Siemens Brothers & Obach, and 115 of the Schuckert system will be required. We have left out 112 ELECTROLYSIS OF WATER. the small differences in the operating voltage, since we wish to make only a rough approximation. Further, in making an approximation of the cost of plant we must take into consideration the fact that except when very large plants are to be erected in which a sufficient num- ber of anodes can be placed in series, and connected to an or- dinary lighting circuit, the single pole apparatus otherwise re- quires the installation of a particular kind of low voltage and relatively expensive dynamo machine with all its accessories. COST OF OPERATION. The following costs of operation have already been given for the various processes used in industrial practice, conclu- sions being made on the assumption of either one of the two or the mixed gases being utilized. Along with the cost of production several other factors may be thought of, which, under some conditions, would add materially to the cost of operation. CONSUMPTION OF ANODES. One question would be the using up of anodes when work- ing with iron electrodes in alkaline solutions. The view of most of the constructors of electrolytic apparatus for the de- composition of water that the corrosion of the electrodes may be practically neglected because of the " passive condition " of the iron, appears not to be completely substantiated. This question was discussed in connection with the presentation of the Schmidt apparatus at the seventh yearly Assembly of the German Electrochemical Society of Zurich. According to the communication which Heraeus has made upon the basis of the operation of his Schuckert plant, the consumption of anodes is not unimportant. Heraeus supposed the cause to be primarily in the presence of chlorine and sulphuric acid in the cheap caustic soda solution, and replaced the same by the purer and more costly caustic potash, without, however, dimin- ishing the consumption of the iron electrodes. The electrodes lasted in the latter case as in the former about one year. The APPLICATIONS. X. Cost of Producing the Gases per Cubic Meter. System. Mixed gas. Hydrogen. Oxygen. Remarks. Power costing Power costing Power costing Kc. kilo- watt hour. Cents. I#C. per kilo- watt hour. Cents. *c. per kilo- watt hour. Cents. i#c. per kilo- watt hour. Cents. #c. per kilo- watt hour. Cents. i*c. per kilo- watt hour. Cents. A. Pure cost of power for the electrolysis without compression. i Schmidt ... I.OOO 0.900 1.540 1.050 1.225 1.325 5.000 4.500 7-750 5-250 6.000 6.750 1.500 1.325 2.325 1.540 1.850 2 025 7.500 6.750 1.150 7-750 9.250 1.025 3.000 2.750 4-500 3.000 3-675 4-075 15.000 13-500 23.250 15-500 18.375 20.400 2. Schoop : (a) alkaline (b) acid 4. Siemens & Obach 5- B. Total cost of produc- tion including sink- ing fund and inter- est, but without com- pression. i Schmidt 7-775 4.000 5.000 3-75 13-250 11.500 9.250 11.250 5.750 7-500 5-500 19.250 15.000 13-500 22.000 11.750 15.000 II.OOO 38.000 1 30-500 27.250 Only power and sink- ing fund and interest only on the electro- lytic plant. Power without gen- eral expenses. Without sinking fund and interest on the cost of gasometers, as well as without general expenses. 2. Schoop (acid) .... C. Total cost of produc- tion including sink- ing fund, interest, and compression of the gases. 28.750 5-650 35-5oo 38.250 8.500 1 47.000 68.750 14.500 86.500 Without interest and sinking fund on the investment in flasks as well as general expenses. electrolytic formation of ferrates may also indeed be assumed to take place, which explanation Haber 1 has fixed upon re- cently as the result of his extensive study of the subject. Schmidt says, on the contrary, that after using the apparatus of his system two and a half years, the electrodes had only di- minished i mm. in thickness. It is, however, to be remembered 1 Zeitschr. f. Elektrochemie, 1900-1901. 114 ELECTROLYSIS OF WATER. that Schmidt does not use caustic alkali, but potassium car- bonate, which is very free from sulphuric acid and chlorine. ABSORPTION OF CARBON DIOXIDE. A further alteration may take place in the charge of the electrolytic apparatus for the decomposition of water, working with caustic alkaline electrolytes, which is due in part to the absorption of carbon dioxide from the atmosphere. On this point it is to be remembered that the most modern apparatus prevent as far as possible the contact of air with the electro- lyte, in which a layer of mineral oil upon the surface of the electrolyte gives good service, and that finally with apparatus working at high temperatures a layer of vapor above the elec- trolyte excludes the carbon dioxide of the atmosphere in a satisfactory manner. SECURITY FROM EXPLOSIONS. Concerning the safety of the gases from explosion, the limit of safety with oxygen lies at 90-95 per cent, purity. All of the newer apparatus are completely above the requirements in this respect. CONCURRENT PROCESSES. Three groups of processes may be brought out as competi- tors of the electrolytic decomposition of water, namely : (1) Other electrochemical processes in which oxygen or hy- drogen are obtained as by-products. (2) Physical processes. (3) Purely chemical processes. (0 Electrochemical Processes. (a) Hydrogen. Hydrogen is the most important of the by-products of elec- trochemical processes. All those processes which rest upon the electrolytic decomposition of the alkaline chlorides evolve this gas in important amounts. Of these processes, those which work without a separation of the electrode spaces, like the manufacture of chlorates and hypochlorites (usually called ELECTROCHEMICAL PROCESSES. 115 electric bleach), hardly permit of the separate utilization of the hydrogen, since, on the one hand, especially with electric bleach, the installation is mostly too small; on the other hand, with both classes of processes, the hydrogen is usually con- taminated in not unimportant quantities with the products set free at the anode (O, Cl, etc.). In those processes, on the other hand, which decompose the alkaline chlorides for the purpose of manufacturing separately chlorine and caustic alkali, hydrogen can be obtained in the pure state. In such methods of working, under this heading, which use diaphragms for instance (processes of the Elektron companies), the catch- ing of the hydrogen will entail not very convenient altera- tions in the apparatus. Such methods have, as a rule, at their best, the cathode spaces only fairly tight. To enclose tightly the cathode spaces would be quite incon- venient when the diaphragms have to be changed. The pos- sibility and the practicability of obtaining hydrogen as a by- product has been already proven. As an example, the Ver- einigten Chemischen Fabriken Leopold shall sold compressed hydrogen which had been obtained as a by-product, but its production has since been given up. The obtaining of hydrogen as a by-product is the easiest with those electrolytic alkali processes which work with mer- cury cathodes. The amalgam here lies in a space distinct from the decomposing apparatus, and so only this space need be closed air-tight and provided with suitable piping. This opportunity brings to the author's remembrance a very interesting application of the hydrogen set free by the de- composition of this amalgam. He had at one time the occa- sion to inspect and report on an experimental plant at Aachen run on the Stormer system. Since the utilization of the hydrogen was not to be thought of on account of the small- ness of the plant, Stormer used the same for an empirical con- trol of the efficiency of the apparatus. The hydrogen escaped from the tightly closed amalgam washing apparatus by a nar- row vertical tube, was ignited and the height of the flame gave an approximate idea of the efficiency. Il6 ELECTROLYSIS OP WATER. Aside from the manufacture of chlorates and electric bleach, omitted for the above explained reasons, we may assume the power used in the various plants which are concerned with the electrolytical decomposition of alkaline chlorides at in round numbers 45,000 kilowatts. If we assume an average current output for all competing processes at 80 per cent., and the average working voltage of 4.5, there results a daily pro- duction of hydrogen of 80,000 cubic metres, the greater part of which at present escapes unused. There is therefore no prospect of success for the electrolytic manufacturer of hydrogen when increasing consumption of the chlorine and alkali plants permits them to bring their by- product, hydrogen, in a compressed state on the market. Those cases are naturally excepted which are so placed with regard to a hydrogen market that the transportation of the compressed product is too high on account of freight. () Oxygen. Other electrochemical processes scarcely enter as competi- tors for the manufacture of electrolytic oxygen. There is in- deed a possibility that oxygen may be produced as a by- product in electrolytic processes working with insoluble anodes. This will, however, in most cases be obtained with hydrogen according to the better or poorer efficiency of the cathodic precipitation. (2) Physical Processes. (a) Oxygen. In the line of physical processes, the Linde method of ob- taining fluid air is a competitor with which the electrolytic manufacture has to reckon. According to Linde's latest com- munication 1 , he is in a position to produce fluid air with an expenditure of 3 horse-power hours per kilogram with his smaller machines or with his largest type with 2 horse-power hours per kilogram, and believes it possible to reduce the power requirement to i ^ horse-power hour per kilogram. 1 Zeitschr. des Vereins deutscher Ing., 1900, 70. PHYSICAL PROCESSES. 117 CO SO 40 30 20 *> 0% Fig. 85. He reckons the total cost of liquid air with large plants at 2 J^ cents per kilogram. These assumptions do not take into account the large losses which fluid air suffers in transporta- tion. The possibility of fractioning fluid air into gas mixtures rich in nitrogen and oxygen, respectively, is well known. When the separation takes place at atmospheric pressure the re- sults are as shown in Fig. 85, as given by Linde. At the be. ginning of the evaporation, the gas mixture consists of about 92 per cent, nitrogen and 8 per cent, oxygen. This relation changes according to the curve a c b, on different evaporations. The curve d e shows the en- riching of the respective values of oxygen. If an electrolytic plant utilizes hydrogen as well as oxygen, it can compete with the Linde process accord- ing to the costs of manufacture so far given, assuming similar cost prices for power, because the electrolysis requires a smaller investment in plant and, therefore, a smaller sinking- fund and interest cost. The conditions are not so favorable if the hydrogen cannot be utilized. The figures would be still further to the advantage of Linde if the hopes of this inventor materialize according to which, by utilizing a new principle of construction, he expects to make one cubic metre of gas containing 50 per cent, of oxygen with an expenditure of i horse-power hour. In such a case, Linde calculates a total cost of i cubic metre of gas containing 50 per cent, of oxygen at 0.625 cent > which is equal to 0.3 cent per horse- power hour. The electrolytic oxygen would be able to compete, without doubt, in all those cases in which pure gas must be had, since fluid air under the best conditions furnishes a mixture contain- ing only 70-75 per cent, oxygen. The electrolytic method Il8 ELECTROLYSIS OF WATER. would, also, according to all probabilities hold the field for small installations. (3) Chemical. (a) Hydrogen. The purely chemical methods for obtaining hydrogen in- cludes solution of metals in acids (iron, zinc, etc.). The elec- trolytic manufacture is technically and economically superior to these processes as will be later shown more in detail. () Oxygen. The chemical process of manufacturing oxygen which rests upon the conversion of the oxygen of the air into chemical compounds which easily give up their oxygen, such as the processes based on barium superoxide, normal plumbates, so- dium manganates, etc., furnish only an impure product with about 85 to 90 per cent, oxygen, and what has been said of Linde's fluid air in this respect applies also to them. The processes for manufacturing oxygen from compounds rich in oxygen such as mercuric oxide, potassium chlorate, manganese dioxide, potassium chromate, etc., are not really at present active competitors of the electrolytic process. Compression. Concerning the compression of the gases, it must be re- membered that it is not usually carried over 100 to no atmos- pheres and only in extreme cases are pressures of 200 atmos- pheres used. Schoop gives some details of compression plants in his recently published treatises. 1 The steel flasks are usually made to contain from 10 to 250 litres. The weight is, on the average, 10 kilograms to a cubic metre of gas. The most practicable sizes of flasks are shown in the following table taken from Schoop's work : 1 M. U. Schoop : Die industrielle Elektrolyse des Wassers, 1901. SPECIAL APPLICATIONS. 119 TABLE XI. Length in metres. Outer diameter in metres. Approximate weight in kilo- grams. Water in litres. Price of the empty flasks in dollars. 0.432 0.076 6 1.40 6.00 0.609 0.102 II 3.68 6.75 O.6OO 0.140 20 7.00 7.50 0.930 0.140 20 11.00 8.25 1.346 0.140 40 16.70 11.75 I.O9O 0.203 45 26.80 15.00 2.OOO 0.205 73 50.00 20.00 Some producers of oxygen and hydrogen in order to render certain the recognition of the oxygen and hydrogen flasks from each other, provide the oxygen flasks with right-handed valves and the hydrogen "flasks with left-handed valves and black nozzles. Special Applications. The uses of oxygen and hydrogen and the mixed gases are generally known and will here be collected as a conclusion of this Monograph, and some new propositions concerning their use alluded to more at length. (i) Detonating Gas. (0) HIGH TEMPERATURE. For high temperatures, in particular the working of metals : This application dates back to the work of St. Claire- Deville who melted platinum in the oxyhydrogen flame. In many cases the electric furnace can be replaced by the oxy- hydrogen flame and in this connection we may recall the attempts to manufacture calcium carbide without the electric current. More recently the question of using mixed gas in glass furnaces has regained interest. An investigation of Ascherl has been in this direction. The manufacture of large glass vessels from plates of glass, melting down the joints with the oxyhydrogen flame, is carried out in England on an indus- trial scale. The oxyhydrogen flame has gained an important place in many cases in the working of metals where the nature of the 120 ELECTROLYSIS OF WATER. metal is such as to exclude the use of carbonaceous gas even when mixed with oxygen. A content of carbon either hinders in most cases the soldering entirely, or deteriorates the quality of the metal soldered. Repairing of water-tube boilers can be easily accomplished with an oxyhydrogen flame. We shall speak of the soldering of lead further on when treating of the applications of hydrogen, since mixed gas is not used for this purpose but a mixture of hydrogen and oxygen poorer in the latter. () LIGHTING. Lighting : Mixed gas has been used for this purpose for a long time for the so-called Drummond lime light, and has attained, for special purposes, a certain sphere of usefulness, although a limited one. In the recent Spanish-American war it was used on a large scale for search-lights. Mixed gas never was able to hold a place for general light- ing purposes in spite of the stirring activity which many in- vestigators have shown in this direction. The principal diffi- culty is indeed in the great danger of explosion, and the difficulty of ob- taining durable incandescent mate- rial. Investigation in this direction has not yet been given up. For in- stance, there lies before the author a recently allowed Austrian patent of the 1 5th of May, 1901, to Urban- itzky. According to this, the electrolysis of the water takes place at each lamp and the mixed gas makes an incan- descent substance glow. In the drawing, shown in Fig. 86, a b is the Fig 86 decomposition vessel, c d the elec- trodes, ef the gas-exit tubes which end in the double tube g, h the glowing body, and i the igniting device for the mixed gas. SPECIAL APPLICATIONS. 121 Quite aside from the conditions and considerations of tech- nical operation, the patentee seems to have quite forgotten what sized section of conductors he would require to conduct to each single lamp the necessary current for the decom- position of water at the low tension of 2.5 to 3 volts. (V) BLASTING PURPOSES. Blasting purposes. The attempts to use mixed gas for blasting purposes are not so generally known. Siemens & Halske, A. G., in Vienna, took up experiments in 1893 to a Pply electrolysis directly to blasting purposes. The first experiment was to produce haloid nitrogen com- pounds in small, thick capsules at the place where it was to be ignited. Later the investigations were extended to com- P * Utv Fig. 87. Fig. 88. pressed oxyhydrogen gas. The decomposing apparatus were thick-walled, pear-shaped, glass flasks with two platinum wires sealed therein (Fig. 87). These were filled with the electrolyte so far that the pla- tinum wires were not entirely submerged and then sealed. The ignition was produced after the electrolysis had been finished by connecting the poles with an induction coil. Larger experiments were made with the apparatus shown in Fig. 88. The affair was not further investigated because, primarily, of difficulties with the patent. 122 ELECTROLYSIS OF WATER. Ochse tried to utilize the same idea. 1 He used steel cylin- ders, 1 80 mm. long, closed by a stopper which carries the electrodes and the igniting apparatus. These cylinders can withstand a pressure of 1200 atmospheres and contain 22.5 grams of water and 2.5 grains of soda solution. Using 8 to 10 volts tension, a current strength of 0.8 to i ampere was obtained. The electrolysis was continued until 20 grams of water were decomposed and then the explosion of the mixed gas produced by an induction spark. The explosive power of such a cartridge is stated to be equivalent to 150 grams of modern safety explosive containing ammonium nitrate (robu- rite, bellite, securite, etc.). Cornara has embodied the same idea in the English patent, 302,53 of the 2ist of December, 1897, except that the shape of the cartridge was somewhat modified. More recently it appears that investiga- tors have given up the idea of generating and compressing the mixed gas in the cartridge itself. The attempts to use mixed gas for blasting purposes have taken more the direction of manufacturing the mix- ture of gases outside the cylinder and then compressing it mechanically into the latter. As an example of such a cartridge, we illustrate in Fig. 89 the Boehm filling ar- rangement for oxy hydrogen-blasting cart- ridges patented in Germany 107,531. In this : 1. The cartridge case. 2. The screwed in closing head. 3. The joint stopper. 4. The igniting electrode. 5. The steel tube for filling the cartridge and likewise acting as the second electrode. 6. The igniting filament. 7. The screw attachment for fastening to the air-pump. 8. The screw for adjusting the valve stem. 9. The valve stem. TO. The cone joint. 1 D. R. P., 67153. Jacobsen, Repertorium 1893., I, 238. El. Engineer, London, XXVII, No. 19, 13 May, 1898. Electrochem. Ztschr., 1898-99., 127. SPECIAL APPLICATIONS. 123 We cannot so far speak of the definite introduction of oxy- hydrogen blasting, either by the direct or indirect production of the gases. (2) Hydrogen. (a) BALLOONING PURPOSES. Ballooning: For this purpose hydrogen has been used for a long time, and the electrolytic production of hydrogen has by far surpassed all other methods of producing it. For the purposes of ballooning, the hydrogen should be as pure as possible in order that the lifting power of the balloon may be the greatest possible with the smallest possible cubical content. By doing so, on the one hand, expensive balloon material is spared ; on the other hand, the cost of transporta- tion of the compressed gas is less. It may be observed that hydrogen obtained in the chemical way from iron and sul- phuric acid is nearly double as heavy as the pure gas (160 : 89), while electrolytic hydrogen can be obtained easily only about 25 per cent, heavier than the pure gas. 1 Up to the present, the Italian, French and Swiss army bureaus manufacture their hydrogen for balloon purposes electrolytically. The German army uses compressed by-product hydrogen from the electro- lytic alkaline chloride plants. () SOLDERING PURPOSES. Soldering purposes : In this line the most prominent applications are in the manufacture of accumulators, and in the sulphuric acid works for the purpose of soldering lead. Clean lead surfaces, it is well known, may be easily soldered with a blowpipe with a hydrogen-air flame. The hydrogen necessary for this was formerly exclusively manu- factured by the use of crude zinc and dilute sulphuric acid (15 and 20 per cent). It would lead us here too far to dis- cuss the more usual chemical apparatus employed for evolving this hydrogen and so we refer as far as concerns it to the litera- ture of this subject. 2 1 W. Diirer : Electrochemische Zeitschr., VIII, 2 (1901). 2 Z. B. P. Schoop: Die Sekundarelemente, II, 41 (1895). Zeitschr. f. Elektrochemie, II, 203 (1895-96). 124 ELECTROLYSIS OF WATER. One of the principal disadvantages of the chemical process lies in the circumstance that in dissolving the zinc in sul- phuric acid the arsenic in the latter is evolved as hydrogen arsenide according to the equation As 2 O 3 + 6H 2 SO 4 + 6Zn = 2AsH 3 + 3H 2 O + 6ZnSO 4 . A content of 0.5 per cent, of arsenic is not unfrequently found in sulphuric acid and it is, therefore, easily seen that constant working under such conditions with hydro- gen generators, which are often not perfectly tight and there- by allow unburned gas to escape from burners which are extinguished but not entirely turned off, results in the most disagreeable sanitary consequences for the workmen. For this reason the accumulator constructors themselves have pro- posed to pass this chemically prepared hydrogen through red hot copper tubes and afterwards to cool it, by which operation the arsenic is seperated out, or the gas is freed from arsenic by washing with potassium permanganate solution. 1 These sanitary difficulties are naturally completely avoided by the use of electrolytic hydrogen. Also the danger of ex- plosion in the apparatus itself is less with this manner of manufacture. Finally when using electrolytic hydrogen gen- erators, the oxygen simultaneously generated can be used in place of air in the blast-flame, which results in a saving of about 50 per cent, of the time necessary for soldering. Also from the point of view of economy the electrolytic method of manufacture has the advantage over the chemical. For i kilogram of zinc costing, to-day, 8^ cents, 2 kilo- grams of sulphuric acid are required, costing 3)^ cents. For this outlay of 12 ^ cents there is theoretically only 32 grams or not quite 360 litres of hydrogen to be obtained. Since the zinc and acid are not completely used up, it may be assumed that 800 ampere-hours would electrolytically produce an equal volume of hydrogen. Using 2.5 volts working tension, 2 kilowatt hours would be necessary, which, at the price of 1 P. Schoop : Die Sekundarelemente, II, 45, (1895). SPECIAL APPLICATIONS. 125 power already assumed ^ to i J^ cents, would cost y 2 to 2j4 cents. The total costs for the soldering have been determined by the results of operation in various accumulator factories to be about one-half cheaper with the electrolytic apparatus than with the chemically generated hydrogen. Besides the econ- omy in the cost of gas and time, there can also be taken into consideration the diminution of the labor required. The gas must naturally be compressed for use outside, yet the weights to be transported in gas flasks are no greater in the one case than in the other. M. U. Schoop has recently published an interesting disser- tation on the soldering of lead with compressed oxygen and hydrogen. 1 Since this, however, does not include the electro- lytic production of the gases, but only their use after being compressed and their extensive applications in this direction, we refer to the original paper. It may be, however, mentioned that since the soldering flame must have reducing qualities, when electrolytically pro- duced gas is used, there is a certain excess of unused oxygen. The gases are consumed in the proportion of i hydrogen to j oxygen instead of i hydrogen to y 2 oxygen as furnished by the electrolyzers. Soldering with the soldering iron and easily fusible solder often competes with the hydrogen soldering, yet according to the reports of professional people a cleaner lead surface is necessary and the quality of the soldered joint is not equal to that produced by the autogenous soldering of the lead, quite aside from the considerations of the sanitary inconveniences resulting from the content of mercury in the easily fusible solder. In the sulphuric acid industry the hydrogen soldering with electrolytically generated gas is of particular importance be- cause of the quick soldering of vertical seams. Heraeus in Hanau has used with good success the blast- flame with a large excess of hydrogen for the autogenous 1 Zeitschr. f. Elektrotechnik, Wien, XIX 224, (1901). 126 ELECTROLYSIS OF WATER. soldering of aluminium. The samples of work shown at the Paris Exposition attracted considerable attention. (V) LIGHTING. Lighting: The attempts for using hydrogen for the purposes of lighting go back to the year 1846 when Gillard, in France, introduced a process by which a little basket of platinum wire was brought to a white heat by burning hydro- gen. A successful result was not obtained, as was also the result of attempts of White and Leprince to use carburized hydrogen. More recently Schmidt has taken up the applica- tion of hydrogen for lighting purposes in combination with the Auer (Welsbach) mantle. Schmidt first presented his views upon this subject before the Seventh Yearly Assembly of the German Electrochemical Society in Zurich, in 1900, and a lively debate arose on the subject, in which especially Nernst brought out the high re- quirements of the mantles for the use of hydrogen gas, and Forster the difficult question of suitable stop-cocks. Concerning the views of Schmidt he has most accurately expressed them to the author in a long private communication, the most important of them being here reproduced in abstract : The most advantageous qualities of hydrogen which appear to make it well suited for lighting purposes are its harmless- ness for the human organs and high heat of combustion, the small weight, the relatively high ignition point, the easy conductibility, and the chemical inactivity of the same towards the fittings. Its combustion produces only water vapor, no foul smelling products, and it takes from the atmosphere the smallest amount of oxygen. The flame is really non-luminous but will heat a Welsbach mantle to the brightest, white heat, whereby a mantle of ordinary size will give a far too great amount of light (several hundred candle-power) for or- dinary purposes. Schmidt uses, therefore, in his light investigations, a very much smaller mantle, about 6 cm. long and 0.5 cm. in diameter, one of which is shown in Fig. 90. The smaller di- SPECIAL APPLICATIONS. I2 7 mensions makes the mantle more proof against damage by vibrations. The difficulties named by Nernst of the greater require- ments of a mantle for withstand- ing the hydrogen flame are, ac- cording to the view of Schmidt, balanced by the circumstances that the hydrogen contains no im- purities, as iron compounds and dust which in other cases affect the durability of the mantles. The inconvenient quality of being without odor is shared by both hydrogen and water-gas, and this difficulty in the use of the gas for lighting can be overcome in the case of hydrogen in the same way it is managed with water, i. e., by introducing into it small quanti- tities of mercaptan vapor. A very small amount of hydro- gen gas needs to be used on ac- count of its high temperature of combustion. Schmidt compares in his experiments hydrogen with illu- minating gas and acetylene, and believes the superiority of the first to be proven by the following facts : (1) Sanitary advantage, since only water vapor results as a product of combustion and the minimum amount of oxygen is withdrawn from the atmosphere with a much less develop- ment of heat. (2) A very much smaller section of piping required. These are for hydrogen, used at a similar pressure and volume, one- eighth the size necessary for illuminating gas and one-tenth that for acetylene. Referred to the candle-power developed, the sizes of piping necessary for hydrogen is only one-fifteenth and one-ninth that needed for the other gases. Fig. 90. Burner for hydrogen light. 128 ELECTROLYSIS OF WATER. (3). Very much easier and less dangerous compressibility to many hundred atmospheres without altering the qualities of the gas, while illuminating gas is rendered valueless by com- pression, the oil and fat gas used for lighting cars can be compressed only to 10 atmospheres and acetylene cannot be compressed higher than 2 atmospheres on account of the dan- ger of explosions. (4). The possibility of higher pressures in the gas pipes, so that the flame may be turned in any direction, even down- wards, which is practicable with a pressure of 200 mm. of water. (5). Smaller dangers of explosions since the latter can only take place when it contains 9.5 per cent, of air, on the con- trary with acetylene when it contains 3.8 per cent., with illu- minating gas 8 per cent. ; since the difTusibility of hydrogen is twenty times greater than that of air the mixing of the gas with the air takes place much more rapidly. Concerning the cost of lighting with hydrogen, Schmidt gives the following comparative figures for the above men- tioned kinds of gas. Using illuminating gas with a heat value of 5000 calories when burned in a Welsbach burner, there is used per normal candle-power hour 2.1 litres, if used at 220 mm. pressure i litre, with hydrogen somewhat less than i litre, with acetylene 0.75 litre, and with the proposed but not yet achieved com- bustion of the latter in the Welsbach mantle 0.36 litre. Since the consumption of gas varies naturally with the strength of the flame, Schmidt found the consumption of hydrogen per candle-power hour to be in small burners 1.5 litres for each burner, for large burners i litre, and as an average 1.25 litres. With reference to their production by means of electricity, the present values between the hydrogen light and the calcium carbide, that is, acetylene, is to the advantage of the former. One kilowatt day produces 4 cubic metres of hydrogen and equals 3200 candle-power hours on the one hand, and 4 kilo- SPECIAL APPLICATIONS. 1 29 grams of calcium carbide equivalent to 1200 litres of acetylene or 1600 candle-power hours on the other. In this pre- sentation, however, the very much greater consumption of raw materials and electrodes in the production of calcium carbide in comparison with the electrolysis of water has not been taken into account. In order to attain a more exact basis of comparison, Schmidt starts with the market price of calcium carbide and calculates the same as 6% cents per kilogram, the cost production being 3.625 cents. Out of the latter figures, 1.25 cents is referable to raw materials and its preparation so that 2.375 cents remains for current, furnacing, and general expenses. If the current is reckoned at one cent, there remains for the power only 0.165 cent kilowatt hour, a price only possible in very large plants. Assuming a cost price of 3.625 cents per kilogram of car- bide from which 4oo-candle-power is obtainable, there could be obtained electrolytic hydrogen to a volume of 1540 litres or 1230 candle-power hours. As against this better utilization of power in the manufac- ture of hydrogen must, however, be placed the easier storage and carriage of calcium carbide. Schmidt makes the follow- ing comparisons with reference to the consumption of power for the compression of hydrogen, freight costs for transporta- tion and return of the flasks (10 kilograms per cubic metre), and sinking fund on the assumption of twenty shipments per year. The cost of 10,000 candle-power hours transported 500 kilo- metres (300 miles) : From carbide. From hydrogen. 26 kg. of carbide $ .90 12.5 cubic metres of hydrogen $ .295 Packing .10 Compression .025 Evolution of acetylene, purifica- Interest on flasks .1625 tipn and interest on apparatus .10 Emptying and adjusting valves .0200 Freight 300 miles .125 Freight and return for 300 miles 6250 $1-225 $1-1275 It is seen from this comparison that the distance of 300 130 ELECTROLYSIS OF WATER. miles is about the limit at which hydrogen can compete with acetylene. For small distances, distribution in pipes would naturally be used. What difficulties would in that case be encountered in preserving the tubes gas-tight cannot be necessarily fore- told from an experimental study of the subject. (d) MOTOR PURPOSES. Motor purposes : The evident applicability of hydrogen for motor purposes has likewise been often taken into con- sideration, but at the present no results worth mentioning have been obtained. (3) Oxygen. Concerning, finally, the application of oxygen there is no question that with a sufficiently low price, quite a number of extended applications would be found. The hastening of oxidation processes and intensifying of combustion is in many cases the end which is being actively sought after. In many steel works, electrolytic oxygen is at present being experi- mentally used in place of air in the Bessemer converter. Even in various blast-furnace processes the application of oxygen has been thought of if the difficulties could be over- come in finding a suitable material for constructing the fur- nace. Investigations in the glass industry have shown that the introduction of oxygen into the fluid mass of glass results in an economy of 40 to 50 per cent, in labor, time, fuel and de- terioration, without damaging the quality of the glass. The extensive application of oxygen in the manufacture of sul- phuric acid according to the contact process is indeed only a question of time. In organic technology, oxygen is already used for many purposes. It is used, for instance, partly as ozone for the ageing of alcoholic beverage, in the manufacture of varnishes, and in the purification of illuminating gas. More recently oxygen has been used in the preservation of milk, the freshly SPECIAL APPLICATIONS. 131 obtained milk being saturated with oxygen at a pressure of 5 to 6 atmospheres and allowed to stand several hours under this pressure. After diminishing the pressure to two atmos- pheres, the milk is ready for transportation. In the production of ozone, oxygen would be used in all those cases in place of air, in which the introduction of nitro- gen compounds into the product treated with ozone is to be avoided. Finally compressed oxygen has manifold uses in medicine and hygiene, and in particular for the improvement of the air in rooms where renewal of the air is not possible, also for instance in mines and at fires or under such conditions where very attenuated air makes breathing difficult, as in balloon ascensions. It would surpass the limits of this volume to go into the methods of analysis -of the gases of which we have treated, and concerning this subject we refer to the numerous special treatises. 1 There remains only to the author the agreeable duty to thank most heartily Messrs. Dr. Hammerschmidt, of Niirn- berg, Dr. Schmidt, of Zurich, and M. U. Schoop, of Cologne, for the furnishing of original material, and also Dr. Abel, of Vienna, for friendly help in the reading of the proof. 1 Bunsen : Gasometrische Methoden. W. Hempel : Gasanalytische Me- thoden, II Auflage, 1890. Cl. Winkler : Lehrbuch der technischen Gas- analyse, III Auflage, 1901: B. Neumann: Gasanalyse und Gasvolumetrie, 1901. 1 c ^ O 03 o 7 I XJ O S S rt * 2 s . (X, g OJ rt ^3 H M . 3 H d Q X 55 W I! 9 3 ' = M CM to * IO vO t^OO ON O W ~ vN 5 S^ t -< >-< vO 1^00 ON O >-i CM tO T IO VO r*GO ON "-< t-i CM CM tO f >~; CM' to <* 10 vo r^co' d o' d d d d d d d d <-> CM to - 10 VO t^OD O- O OOVO-^CMO I~lOfO^ o >-! CM to ^- ^ iO\O r^co M CM to < 10 vd t^-co' o^ d lOMVOi-'VO CMt^CMOO 8 q q q q q q q q q M M CM CM to to M' CM' to 4 10 vd t^-co o^ d do d d fOiOt^OtO vOCO'-itO OOO-Hi-4 H-i^CMCMCO q q q q q q q q q q t-J CM to 4 10 vd *>-oo' ON O '^3^? 'S'iSSff'g t^iocMOr^ Tj-cMONt^ o M CM' to -4 iOvd rioo' ON - CM ON t GO t^ t o' M CM to 4 iOvd i>-od o\ 5 tO -00 ) OO t^ t>-vO vO i NC>0^ O-i CM CO * IOVO I--00 O i-i CM CO * IOVO 1^00 i iovO ( ' W - -joqoq'oo-oq O M * CO TJ- lO^O lA rO O\ \00 t^^O LO Tf cO C 1 * -< \ c^ ON o^oo ocoo r jcoot^. "P^ 1 "!^ ^^^^l 1 ,-! ^ c/5 ^- lOvO r-GO* Q\ ON C> ^00 OO 00 GO* CO o w pi to 4 LO CM o" ^SNvS 5 ^ t-M^OO co co >. t--- r^ * CM ONVO O *-I CM CO 4 IOVO i>-GO ON ON ^00* CO CM ONVO to *-i d M CM' to 4 lo^o' t^-oo' ON 6 M CM' to 4 io\o' t>.x ON dNooc^^ r^rir^o'vd )00 t^l d M CM to 4 lovj cMcqiO"-i t--.-rqoco ONDO' so oo' r^ r-i. t^-^d vd *-i CM to * lOvO t-^OQ ON M M CM M .t^- 1 qvqvq CM GO T o d M CM' to 4 iovO r^ CT TT co-*^ vO 1^30 ONO 3R28N8 -> P) ro "* iO VO t^OO O O 00 NO TT CS S'&rg o ci to | W 5^ a i^j a 8 3 A D t % O bo > 3 -8 g S a 1 p H -^ g I S s' B( M P) CO ' o ox) oo t>. t- d >-.' pi co <* IONO" t-^-od o -* co ^ 10 r- 3 p) ex) TJ- q i-i oi co -^ IONO r^-od ON ood 06 od iO O -*r O ^J" 00 COOO vO OO_ O i-i CN! co -^- u^vo' r^-od ON CJ 4->O CC O uo O iO O OC?00< f^l^ O O ON O00_ 00 00 00 t~p J^ d i-i pi co rj- IONO t^-od o NO *-i t^. pt oo coomovo oo t^ u^ -^- cs HH ooo r^ /*} O O ON O ON ONOO CO CO 00 iO >* to P) M ro pi M o O -. -.- -. -. . O ON O OOO O M oi CO * lONO' t^-00 ON CO t^. t^-vO d >-* pi co ^ U")NO r^-od d\ vO t^OO ON O t^.00 O o 10 o 10 o up CN! C [ r>- up m\o r^-oo o CO ONUO O vO HI ONOO r^ \r> UO 'f to N >-< CO CM l-i O O -tJo' ed o O i30 Pi VO O P) P) P) rororr in * ro PJ 11 q ) if invd r*-oo" ON d M pi ro if invd t^-oo' O"N M P) rO Tj- in VO r^OC ON O ONOO t~--vq in T|- ro P< i ci M d *-" pi ro ^ in\o r->*o6 ON vo t^OO ON O f ro pi "' ' iO\O t-~00 ON rovO ~ P4 O\00 6 w' pi ro * ONOO t-^vO ID d ON ONOO t-- t- VO m ONOO i^vo in m ^- ( ONOO f^ t^ ro iO r^ ON 00 t^-vd in e T 06 r^-^o v w N ro M pi rfj-fl- ) ON *} O iO Sr^OrOPl Mt^PloO J-^-iJ-rOP) mcMOPl 8 100 ^ CO 6 >-J ci d *-< pi ro ^ invd t^-oo ON <-> coirvo oo ON PI irjoO -> ON - Tt ro 00 vO m ro w tO t^ 1-4 in ON ONOO 00 t^-vq v O M' pi ro if lOvd l-^-OO' ON O i-" PI ro if mvo t^OO ON O >-i 01 rorf ^ Zf'Q q ^ mvo t^oo ON w P) rO M- m VO r^oo ON O as TABLE XIV. TENSION OF WATER VAPOR IN MM. OF MERCURY FOR TEMPERATURES BE- TWEEN 2 C. to +35 C. (From Regnault's Messungen berechnet von Broch, see Trav. et Mem. du Bur. intern, des Poids et Mes. I. A. page 33, 1881.) c. Ten- sion, mm. c. Ten- sion, mm. c. Ten- sion, mm. C. Ten- sion, mm. c. Ten- sion, mm. 2.0 3-950 + i-7 5.161 +5.3 6.643 + 9- 8.548 +12.7 10.921 9 3-979 1.8 5.198 5-4 6.689 9-i 8.606 12.8 10.993 .8 4.008 + 1.9 5-235 5-5 6.736 9-2 8.664 + 12.9 11.065 7 4.038 5-6 6.782 9-3 8.722 .6 4.067 + 2.0 5.272 5-7 6.829 9-4 8.781 + I3-0 H-I37 5 4.097 2.1 5.309 5-8 6.876 9-5 8.840 I3- 1 II.2IO 4 4.127 2.2 5-347 +5-9 6.924 9-6 8.8 99 13.2 11.283 3 4-157 2-3 5.385 9-7 8-959 13.3 II 356 .2 4.188 2.4 5-424 +6.0 6.971 9.8 9.019 13.4 11.430 r 4.218 2.5 5.462 6.1 7.O2O + 9-9 9.079 13.5 II.505 2.6 5-501 6.2 7.068 13.6 11.580 1.0 4.249 2-7 5-540 6.3 7.II7 + 10 9-140 13.7 H.655 0.9 4.280 2.8 5-579 6.4 7.166 10. 1 9.201 13.8 11.731 0.8 4-3 12 +2.9 5.618 6-5 7-215 10.2 9.262 +13.9 11.807 0.7 4-343 6.6 7.265 10.3 9.324 0.6 4-375 +3.0 5.658 6.7 7-314 10.4 9.386 +14.0 11.884 0.5 4.406 3.1 5.698 6.8 7o 6 5 10.5 9-449 14.1 11.960 0.4 4.438 3.2 5.738 +6.9 7.415 10.6 9-512 14.2 12.038 0-3 4.471 3.3 5-779 10.7 9-575 14.3 12.116 0.2 4.503 3.4 5-820 + 7-0 7.466 10.8 9.639 14.4 12.194 O.I 4.536 3.5 5.860 7.1 7.517 +10.9 9-703 14.5 12.273 3.6 5.902 7.2 7-568 14.6 12.352 0.0 4.569 3-7 5.943 7-3 7.620 + 11. 9-767 14.7 12.432 +0.1 4.602 3-8 5.985 7.4 7.672 ii. i 9.832 14.8 12.512 0.2 4.635 +3-9 6.027 7-5 7-725 II. 2 9.897 +14.9 12.593 0-3 4.668 7-6 7-777 H.3 9.962 0.4 4.702 +4.0 6.069 7-7 7.830 II.4 10.028 +15.0 12.674 0-5 4.736 4.1 6. 112 7.8 7-883 II.5 10.095 15.1 12.755 0.6 4.770 4.2 6.155 +7-9 7-937 u.6 10.161 15.2 12.837 0.7 4-805 4-3 6.198 ii. 7 10.228 J5-3 12.920 0.8 4.839 4.4 6.241 +8.0 7.991 ii. 8 10.296 15.4 13.003 +0.9 4.874 4-5 6.285 8.1 8.045 + 11. 9 10.364 15-5 13.086 4.6 6.328 8.2 8.100 15-6 13.170 +1.0 4.909 4.7 6.373 8-3 8.155 + 12. 10.432 15-7 13-254 i.i 4-944 4-8 6.417 8.4 8.210 12. 1 10.501 15-8 13-339 1.2 4.980 +4-9 6.462 8.5 8.266 12.2 10.570 4-^5-9 13-424 i-3 5.016 8.6 8.321 12.3 10.639 1.4 5-052 +5-0 6.507 8.7 8.378 12.4 10.709 + 16.0 13-510 1.5 5.088 5-1 6.552 8.8 8.434 12-5 10.780 16.1 13-596 + r.6 5-124' +5-2 6.597 +8.9 8.491 + 12.6 10.850 + 16.2 13-683 TABLE XIV. (Continued.} c. Ten- sion, mm. c. Ten- sion, mm. c. Ten- sion, mm. c. Ten- sion, mm. c. Ten- sion, n +16.3 I3-770: + 20.1 17.471 +23.8 21.888 +27.5 27.258 +31-2 33-749 16.4 13.858 20.2 17-579 +23-9 22.O2O 27.6 27.418 31-3 33-942 16.5 I3-946 20.3 17.688 27.7 27.578 3i.4 34-136 16.6 14.035 20.4 17-797 +24.0 22.152 27.8 27.740 3i-5 34.330 16.7 14.124 20.5 17.907 24.1 22.286 +27.9 27.902 31.6 34.526 16.8 14.2141 20. 6 18.018 24.2 22.420 31-7 34.722 + 16.9 14-304 ; 20.7 18.129 24-3 22-554 31-8 34.920 20.8 18.241 24-4 22.690 +28.0 28.065 +3r-9 35-119 +17.0 14-395 + 2O.Q 18.353 24.5 22.826 28.1 28.229 17.1 14.486- 24.6 22.963 28.2 28.394 17.2 14.5/81 24-7 23.100 28.3 28.560 +32.0 35.318 17-3 14.670 ! + 21.0 18.466 24.8 23.239 28.4 28.726 32.1 35-5I9 17.4 14-763 21. 1 18.580 +24-9 23.378 28.5 28.894 32.2 35-720 17-5 14.8561 21.2 18.694 28.6 29.062 32-3 35-923 17.6 14-950 21.3 18.808 28.7 29.231 32-4 36.126 17-7 15.044 21.4 18.924 +25-0 23.517 28.8 29.401 32.5 36.331 17.8 15.136 21-5 19.040 25-1 23-658 +28.9 29-572 32.6 36.536 +17-9 15-234 21.6 19.157 25.2 23.799 32.7 36.743 21.7 19.274 25.3 23.941 +29.0 29-744 32.8 36.950 + 18.0 15-330 21.8 19.392 25-4 24.084 29.1 29.916 +32.9 37.159 18.1 15-427 +21.9 19.510 25.5 24.227 29.2 30.090 18.2 15-524 25.6 24.371 29.3 30.264 +33-0 37-369 18.3 15.621 + 22.0 19.630 25-7 24.516 29.4 30.440 33- i 37.580 18.4 J5-7I9 22.1 I9.750 25-8 24.662 29.5 30.616 33- 2 37.791 18.5 15.818 22.2 19.870 +25-9 24.808 29.6 30.793 33-3 38.004 18.6 I5.9I7 22.3 19.991 29.7 30.971 33-4 38.218 18.7 16.017 22.4 2O.II3 +26.0 24.956 29.8 3 I - I 49 33-5 38.433 18.8 16.117 22.5 20.236 26.1 25.104 +29.9 3I-329 33-6 38.649 + 18.9 16.218 22.6 20-359 26.2 25.252 33-7 38.866 22.7 20.482 26.3 25.402 +30.0 31-510 33-8 39.084 +19.0 16.319 22.8 20.607 26.4 25.552 30.1 31.691 +33-9 39.303 19.1 16.421 + 22.9 20.732 26.5 25.703 30.2 3L873 19.2 16.523 26.6 25.855 30.3 32.057 +34-0 39.523 J9-3 16.626 26.7 26.008 30.4 32.241 34-1 39-744 19.4 16.730 +23.0 20.858 26.8 26.l6l 30.5 32.426 34.2 39.966 19-5 16.834 23.1 20.984 +26.9 26.316 30.6 32.612 34-3 40.190 19.6 16.939 23.2 21. Ill 30.7 32.800 34-4 40.414 19.7 19.8 17.044 17.150 23.3 234 21.239 21.367 +27.0 27.2 26.470 26.626 30.8 +30.9 32.988 33.176 34-5 34-^ 40.640 40.866 + 19-9 17-256 23.5 21.496 27.2 26.783 34-7 41.094 23-6 21.626 27-3 26.940 +31.0 33.366 34-8 4i-3 2 3 +20.0 17.363 +23-7 21.757 +27-4 27.099 +31.1 33-557 H-34-9 4L553 TABLE XV. CONDUCTIVITY OF THE ELECTROLYTES WHICH ARE MADE USE OF IN THE TECHNICAL ELECTROLYSIS OF WATER. The table is for the temperatures of the solutions of 18 C. P = Percentage by weight of the anhydrous electrolyte in 100 parts of the solution. t\ = Number of gram-equivalents in i cc. of the solution ; in the calcula- tion of the gram-equivalents and litres m 1000, n = the concentra- tion, or v = the dilution. m s = Specific gravity of the solution between 15 and 18 referred to water at 4 C. XK Conductivity in at 18 C. The temperature coefficient gives in terms of x^ the alteration of x for +1, and indicates the mean change between 18 and 26. Interpolated values are enclosed in brackets. Electrolyte. p. IOOO TJ stk. 10* X W . (m; ilv) "i 18 V