CHARTOGRAPHY 6 4 8 7 ARYt ACUITY 1 1 TEN^'lJ^S^SOiNS rnANR J.VWARNE CHARTOGRAPHY IN TEN LESSONS BY FRANK J. WARNE AUTHOH OP warne's book of charts warne's elementary course in chartography, etc. ILLUSTRATIONS BY H. F. CHURCH PUBLIHHED BY FRANK J. WARNE, SOUTHERN RUILDING WASHINOTON. r). C. Copyright, 1919, By F. J. WARNE (Authorship also protected in Great Britain and her Colonies, including Canada) HA^ ■; TABLE OF CONTENTS S Page u. Preface VU Introduction 13C Lesson I '-» Building the Chart •^ Value of Table of Contents 1 (s Statistics and Chartography 3 ''^'^ Definition of a Chart 3 The Statistics 4 Horizontal Lines 5 i| Vertical Lines " i - A Choice of Methods 8 The Use of Pencil Dots 10 The Framework 10 i Lesson II ^ The Scales '^ Statistical Variables 13 The Independent Variable 13 The Dependent Variable 14 The IIorizontHl Scale 15 Detcrrainirif,' the Vertical Scale 15 The Scale Lines Ifi Reading the Scales 1 ' Plotting the StiiHstics 19 The Use of (iradicnlcs 20 iii iv Chartography in Ten Lessons Lesson III The Curve Chart Page Making the Curve Heavier 25 The Horizontal Scale Unit 26 Squares Should be Equal 27 Effects of Different Scale Units 28 Dropping the Zero Line 30 Indicating the Absence of the Zero Line 31 Divisors for the Vertical Scale 33 Lesson IV Features of a Complete Chart The Statistical Table 37 Table Should Appear on Chart 39 The Make-Up of the Table 40 Spacing the Columns 41 The Form of the Table 42 Duplicating the Scale Units 44 The Place for the Horizontal Scale 44 Word Designation of the Scale 46 The Title 47 The Foot-Notes 48 The Neat Lines 49 Lesson V The Bar Chart Making Bars from a Curve 53 Making a Curve from Bars 55 Advantages of the Horizontal Bar 56 Reversing the Scales 57 Width of the Bar 59 Table of Contents v Paob Separation of the Bars 60 Location of the Table 61 The Bar and the Curve 62 Lesson VI The Tools of the Chartographer Cross Section Paper 65 The Lead Pencil 69 The Kind of Ink 70 The Ruling Pen 71 Correct Position for Holding Pen 71 Pen Points 73 The Drawing Board 74 The T-Square 75 The Triangle 75 The Engineer's Scale 77 The Dividers 78 The Essential Tools 79 Lesson VII Accuracy in Charlography The Use of the Typewriter 81 Drawing Letters for the Title 83 Exaggerating the Curve 86 Effects of Exaggerating the Curve 88 Advantages of Extra Squares 91 Lesson VIII Curve and Bar Designations Disadvantages of the Unbroken Curve 93 Curve Designations 95 vi Chartography in Ten Lessons Page Word Designations of Curves 98 The Peak-Top Curve 99 Determining the Scale Spacing 101 Utility of the Curve Chart 102 Chartography Based on Comparisons 103 Bar Designations 105 Interpreting the Bar 108 Some Characteristics of a Good Bar Chart .... 109 Word Designation of Scale Units Ill Lesson IX Value of Statistics to Chartography The Statistical Table II4 Aids in Reading the Table 116 The Substitution of Ciphers II9 The Table of Ratios 120 Building up a Table 12l The Percentage Increase and Decrease 123 The Zero Line 126 The Arithmetic Average 129 The Misuse of the Average 13 1 Statistical Class Limits 132 Lesson X Primary Principles of Chartography Planning the Chart 137 Importance of the Right Method 139 Essentials of Good Chart Making 140 Planning the Size of the Chart 141 Planning a Reduction in Size 143 The Reducing Glass 147 The French Curves 149 Checking up the Completed Chart 151 PREFACE No attempt is made in these ten Lessons to cover the entire field of chartography. To do so adequately would require several large volumes. All that these Lessons aim to do is to familiarize the student with the primary or elementary prin- ciples. These underlie chart plotting ajid con- struction as applied to the curve and bar charts. These principles are applicable to all kinds of charts, variations being explained by the differ- ences in the circumstances surrounding the special problems. These Lessons are the by-product of an experi- ence covering a period of ten years during which time the author has been engaged professionally in the application of the principles of chartography to the working out of practical problems of the work-a-day world. In the presentation by means of diagrams or charts of innumerable statistical problems the author has been able to put the art of chartograj)liy to a severe test before such authoritative tribimals as the Literstate Com- merce Commission, various state public utilities and railroad coinmissions, })oards of arbitration appointed by the President of the United States for the peaceable .settlement of labor contro- vii vdii Chartography in Ten Lessons versies and various committees of Congress. In this professional work he has been called upon to till entirely new fields, for there were few authoritative guides to follow, and in conse- quence he has been compelled to pioneer his way amid innumerable uncharted difficulties. Such an experience should contain lessons of value to others engaged and about to engage in the making of charts. In presenting this formu- lation of the principles of chartography the author begs leave to express the hope that these Lessons may serve the beginner in chartography as a safe guide over the innumerable pitfalls that inevitably must be encountered unless he is warned to guard himself against them. An endeavor has been made to bring these Lessons within a reasonable price to the student. The cost of production of "Warne's Elementary Course in Chartography," which was published two years ago and the price of which was fifty dollars for the twenty Lessons (including "Warne's Book of Charts"), was such as to pro- hibit a lower price, and in consequence it could not be made to serve the class of students the author desired most to reach. Parts of that Course have been completely revised in these Lessons. INTRODUCTION The Value of Chartography (Revised from "Wame's Elementary Course in Chartography" and "Warne's Book of Charts") To the average citizen statistics are as incom- prehensible as a Chinese puzzle. To him they are a mental "Mystic Moorish Maze." He looks upon columns of figures with suspicion because he cannot understand them. Perhaps he has so often been misled by the wrong use of statistics or by the use of incorrect statistics that he has become sceptical of them as repre- senting reliable evidence as to facts and, like an automaton, he mechanically repeats "while figures may not lie, all liars, figure," or the equally common libel, "there are three kinds of lies — lies, damn lies and statistics." And yet statistics are an infallible indicator of economic conditions — they measure the heart- throbs of a nation's or of an industry's life- blood. They register the conditions of any given static situation; they point the direction of a trend or tendency with the accuracy of the ther- mometer in measuring the temj)erature. 1*.^'); Every business, whether organized on a large or a small scale needs statistics — in fact, statistics ix X Chartography in Ten Lessons are vital to its successful existence. Without them the executives cannot know the status or tendency of the economic factors which control their affairs. This practical value of statistics in every-day business life is coming to be more and more correctly appraised at its true worth. The Railway Age of February 23, 1917, says: An officer of a western road recently made the statement that each department of a large railroad system should have on its staff a thor- oughly competent man whose duty it should be to analyze statistics. The assertion was not made without deliberation, and it might be perti- nent to ask whether we are getting the most out of the statistical department. The amounts which are spent by American railroads in com- piling statistics bearing on the functions of the various departments are extremely large, and no one who knows the importance of the compara- tive figures to the oflBcers will question the wisdom of the expenditure. Some of the statistical data which the railroads compile is readily analyzed, and the important figures, such as average tons per train, or pounds of coal per thousand gross ton-miles, are always readily available. A closer analysis of data which is regularly compiled would develop important facts which are not brought out in the routine reports. The head of a department may desire more information than that regularly furnished him, but he cannot take the time to get it himself, and a clerk could not understand its meaning or application, and would Introduction xi overlook important points. Great stress is laid on the comparison of results for successive months, and the comparison with figures for the corres- ponding months of the previous year, but statis- tics are no less important as a means of forecasting results by comparing a proposed method with the one in practice. The great expenditure for statistics is relied upon to show the leaks and determine wasteful methods. The field of the statistician should be broadened and he should give more attention to the possibility of con- structive activity. The statistical department has long been depended upon to keep costs from going up. It is time we recognize that it lies within its province to show how costs can be brought down. Statistics have become a vital, every-day need not only in transportation but also in indus- tries of all kinds, in finance, in journalism, in social work, in public life, and in business of every description. They are necessary to men of affairs, publicists, economists, and even to the average citizen if the significance of facts and the trend of events are to be comprehended. Virtu- ally our entire political life, both state and nation- al, is now regulated and its course determined by statistics. Every important branch of govern- ment has its statistical bureaus. Large financial institutions, industries, manufactories, railroads and other transportation companies have their xii Chartography in Ten Lessons statistical departments. Innumerable associa- tions must resort to statistics in order efficiently and effectively to carry on their work. This recent growth in the demand for reliable statistics and their correct interpretation has suddenly raised the standing of the statistician to one of importance, and this has been accom- panied by a corresponding increase in his remuner- ation. Today the openings for one versed in the fundamental laws of statistics — in their collection, compilation, presentation, and interpretation — are innumerable. The demand is greater than the supply. The value of statistics, while great, is inestim- ably enhanced by the aid of chartography. It supplements statistics — it supplies the best known method for their presentation and interpretation. In all those phases of presentation which count for clearness and quickness of comprehension, no other method is equal to it. It makes clear at a glance, even to the uninformed and unin- itiated, the significance and tendency of the fac- tors that are portrayed. As nine-tenths of the problem in interpretation is clear presentation the value of the service rendered to statistics by chartography cannot be over-emphasized. Especially is this realized when it is remembered that statistics, to be useful and valuable, must Introduction xiii not only be accurately compiled but must also be correctly interpreted. Relatively, too much time is spent in the collection and assembling of statis- tical material and too little in its clear and forcible presentation. Here is where chartography be- comes an invaluable handmaid to statistics. The graphic presentation and interpretation of statistics as the basis of a recognition and an understanding of industrial, political, financial, social, economic, and other tendencies has become essential to practical men of affairs as well as to publicists, economists, and others. It can be made of incalculable value in any line or depart- ment of business — in that of finance, corporate relations, internal organization, traffic, supplies, production, prices, wages, costs, and scores of others. It not only will repay its cost but will be found of such great value in so many ways that, once instituted, it will never be abandoned. It will save in the time of the busy executive alone more than its cost, because it will enable him to analyze the facts and tendencies at a glance in- stead of spending hours in studying the relations of the figures in the different columns. It will make more certain the correct interj)retation of the basal information necessary to action and the formulation of a successful pr)hcy. It will en- large and extend his individual experience and xiv Chartography in Ten Lessons his accumulation of knowledge of the details and principles of his business. It will replace vague- ness and indefiniteness by assurance and certainty, hazy conceptions will become clear-cut perspec- tives, and these in turn will lead to a compre- hensive grasp of the entire problem. The recent development in and the growing demand for diagrammatic statistics will continue, and he who masters the few simple yet funda- mental laws or rules upon which it is based will be in a position to become an authority in his par- ticular field and to command a comfortable in- come. Frank J. Warne. Washington, D. C. October 1, 1919. LESSON I Building the Chart Value of Table of Contents — Statistics and Chartography — Definition of a Chart — The Statistics — Horizontal Lines — Vertical Lines — A Choice of Methods — The Use of Pencil Dots — The Framework. In the Table of Contents preceding this Lesson has been given a comprehensive outHne of the field the beginner in chartography is to cover in this and succeeding pages. It is a bird's-eye view of the course of study that has been mapped out for him in these I^essons. It should not be passed over lightly but should be studied seri- ously, for the reason that such a study will give to the student at the very outset a broad per- spective of the problems he is to encounter and overcome. Figuratively, he is starting on a mental journey, with its ups and downs, its delights and pleasures, its perplexities and obstacles — with a good deal of play and some hard work ahead. Tlie Table of Contents is tlic itinerary, kej)! by one who has many times covered tlie same ground and who thus is able to point out the significance of 1 2 Chartography in Ten Lessons the things that are to be encountered. The student will benefit greatly in the mastery of these Lessons if he will frequently re-read the Contents. VALUE OF TABLE OF CONTENTS The Table of Contents can also be likened to a railroad map in the hands of one starting on a long journey. It enables him to traverse with his eyes the entire distance of the trip, noting the general characteristics of the country through which he is to pass, its mountains and rivers, and the principal cities along the way. Thus he secures, before he starts on the journey, a much better idea of where he is going and becomes more familiar with the country through which he passes than he would if he studied the railroad map piecemeal after the journey begins. These ten Lessons will take the student on an intellectual trip in the course of which he will be called upon to exercise such mental traits as application, concentration, observation, and im- agination in overcoming the various obstacles on his way to the acquisition of knowledge concern- ing the art of chartography. He cannot reach this desirable end without progressing step by step in mastering its various features. And at every step in this progress the broad view of his Building the Chart 3 final destination which he will hav^e acquired by a close and frequent study of the Contents will be of material assistance to him. It will not only enable him to cover the ground much more quickly and with less exertion, but also with much more satisfaction to himself. STATISTICS AND CHARTOGRAPHY The value and usefulness of statistics and the relation to them of chartography, as well as the objects of chartography, have been pointed out in the Introduction. From a reading of those pages it should be plain that figures in tabular form, or which can easily be arranged in the form of a statistical table, are essential to the drawing of a chart — they are the reason for the chart being made. DEFINITION OF A CHART The drawing of a chart therefore presumes the existence of the statistics. It has nothing to do with their collection or compilation. A chart is merely a sheet of paper on which tabu- lated facts are presented grai)hically. It is also called a diagram or "graph." In a limited sense it can l)e likened to a moving pi<-ture, with this difference: In the case of the chart it is the eye and not the picture that moves. 4 Chartography in Ten Lessons the statistics For the purpose of familiarizing the beginner with the various steps in the process of making the framework of a chart these figures are selected : 1913 1914 1915 1916 1917 1918 1919 26.7 26.7 26.4 28.1 38.2 49.5 57.2 The first line of figures represents calendar years and the second line the average retail price of a pound of bacon in the United States on April 15 of each specified year. This information is from page 77 of the Monthly Labor Review of the Bureau of Labor Statistics of the United States Department of Labor. To make a chart from these figures is a simple proposition — as simple as the alphabet, that is, provided one knows the alphabet. It is as difli- cult to one who does not know how as the alphabet is to the child first beginning to lisp the letters. That which at first appears to be a very com- plicated and difficult thing to do comes to be surprisingly simple after one has acquired the necessary knowledge and facility. In the be- ginning all that is needed is a lead pencil, an ordinary ruler, and a blank sheet of paper. horizontal lines A glance at the statistical table shows there Building the Chart ' 5 are seven prices to be recorded. These can be represented for the present by as many lines drawn with the lead pencil at equal distances apart from left to right across the blank sheet of paper. The result gives the lines A-A, B-B, C-C, D-D, E-E, F-F, and G-G on this page. These are horizontal lines. It is important that the begirmer bear this fact in mind. lie should remember that a horizontal line always 6 Chartography in Ten Lessons extends in the direction of the horizon, that is, parallel to the horizon. Here the horizon is represented by the top edge of the sheet. Hori- zontal lines are drawn from left to right, never from right to left. VERTICAL lines In our statistical table we have another set of seven figures. These represent that number of years. So we mark off on the bottom horizontal line G-G, by means of the inch and its fractional units of the ruler, seven dots each an equal dist- ance apart, the first dot starting at the beginning of the line on the left. These dots we repeat on the top horizontal line A-A. Next we draw seven vertical lines connecting these dots, be- ginning with the first dot on horizontal line G-G. Upon completion of the last vertical line erase the dots on the top and bottom horizontal lines. Do not draw these vertical lines backward, that is, downward from line A-A to line G-G. A vertical line is an upright line, that is, it is directed perpendicularly to the plane of the hori- zon, as distinct from the horizontal line which, as has been said, is parallel to the horizon. These seven vertical lines we designate as H-H, I-I, J-J, K-K, L-L, M-M, and N-N. Superim- posed on the seven horizontal lines these vertical Building the Chart 7 lines give the framework shown in the following drawing. K M N H M The distinction between iiorizontal and vertical lines should he clear to the student. The junc- tion of a horizontal and a vertical line forms a right angle. 8 Chartography in Ten Lessons Another way to begin the erection of the frame- work of a chart, and one which will likely appeal more favorably to the student after he has ac- quired greater knowledge of the subject, is to draw first the horizontal lines G-G and A-A and then the vertical lines H-H and N-N. This gives the outline on the opposite page. These four lines are the really important lines of a curve chart. In relative importance they are in this order: H-H, A-A, G-G, and N-N. The uses to which each is put the student will become more familiar with in subsequent Lessons. All that is necessary for him to know now is that: Line H-H is the vertical scale line and with its units of measurement virtually determines the distance the curve is to move. In other words, all movements of the curve are measured by this line. Line A-A is the horizontal scale line, and in all curve charts involving elements of time it takes the time units. In our present problem as to the price of bacon it provides positions for the years. Line G-G is the base line of the chart. Figur- atively, it is the foundation line upon which all the vertical lines rest and from which they start. This base line is the zero of the vertical scale and, whenever possible, should always be indicated Building the Chart 9 by a cipher. All movements of the curve are measured /ro 771 this line. Line N-N is the least important of these four lines, but this is not saying that it is not necessary and useful. Its functions will be pointed out to the student later. 10 Chartography in Ten Lessons the use of pencil dots With the four lines I have described already drawn on the sheet, the beginner next divides by dot markings the base line G-G into six equal spaces, starting the first of the five dots the dis- tance of one space from the left end of the base line G-G and ending the dots the same distance from the right end of the base line. Duplicate these five dots at their respective distance apart on the top line A-A. Now connect these dots with vertical lines extending from the bottom to the top line. This gives the five lines I-I, J-J, K-K, I^L, and M-M. Repeat the pencil dots with the same space between them on vertical lines H-H and N-N, beginning the first of the five dots the distance of one space from the bottom of lines H-H and N-N, Connecting these dots with horizontal lines gives the lines B-B, C-C, D-D, E-E, and F-F. Now erase all the dot markings. The result is the same as that shown on page 7. THE FRAMEWORK OF THE CHART This is the framework of the chart. It is the scaffolding by means of which the curve is to be erected or constructed. It is the skeleton structure for supporting the curve. Without it Building the Chart 11 the curve could not be constructed properly or correctly; neither could the curve adequately perform the service for which a chart is drawn. The lines will be found to occupy positions behind the curve, or rather to form a setting or back- ground for it. The framework is essential for determining the movement of the curve and must be built up before the curve can be placed. QUESTIONS FOR SELF-EXAMINATION 1. Describe the broad view of the field of chartography gained from a study of the Table of Contents. What ser- vice does this table perform for the student? 2. Describe the relation between chartography and statistics. 3. Of what value are statistics to business? To other activities? What is the service chartography performs? 4. Do these Lessons cover the entire field of chartogra- phy? Why? 5. Define a chart. What is a horizontal line? A vertical line? How is each drawn with a pencil? 6. What is the framework of a chart? How is it con- structed? 7. What are the most important lines of the framework? Describe their uses. 8. Of what use are pencil dots in drawing the lines? How are these dots employed in laying out the vertical and horizontal lines? 12 LESSON II The Scales Statistical Variables — The Independent Vari- able — The Dependent Variable — The Horizon- tal Scale — Determining the Vertical Scale — The Scale Lines — Reading the Scales — Plotting the Statistics — The Use of Gradicules. The essence of a chart is in the relation which it shows exists between two or more statistical elements. Chartography involves a comparison. Probably the most frequent comparison is that of figures representing the trend or tendency of the same or similar element or factor over a period of time, as in the present instance of our statistics showing the average price of bacon on April 1.5th of different years. STATISTICAL VARIABLES This price is not the same for all the years — it has the capacity of changing or varying with the different i)eriods of time. Thus in relation to each other these two groups of figures are called variables. THE INDEPENDENT VARIAHLK A comparison being involved, one or the other group must be made use of as the standard by which the other group is measured or interpreted. 13 14 Chartography in Ten Lessons The group so used becomes the independent variable. Where the element of time enters into the situation it is nearly always the standard and thus becomes the independent variable. THE DEPENDENT VARIABLE The statistical group that is to be measured or interpreted is called the dependent variable. In our present problem the price of bacon being dependent upon the specified periods of time is the dependent variable. The relation between or the tendency of the units or elements of the dependent variable is measured by scales. One of these is the horizontal scale and the other the vertical scale. Generally the independent variable takes the horizontal scale. This fact is important, as a great deal of confusion results from a violation of this simple principle of chartography. "It should be a strict rule for all kinds of curve plotting," says Brinton in his Graphic Methods for Presenting Facts, "that the horizontal scale must be used for the independent variable and the vertical scale for the dependent variable. When the curves are plotted by this rule the reader can instantly select a set of conditions from the horizontal scale and read the informa- tion from the vertical scale. If there were no The Scales 15 rule relating to the arrangement of scales for the independent and dependent variables, the reader would never be able to tell whether he should approach a chart from the vertical scale and read the information for the horizontal scale, or the reverse. If charts are always plotted with the independent variable as the horizontal scale, there need be no question in the reader's mind as to how he should interpret the chart." THE HORIZONTAL SCALE Following out this principle of chartography we substitute on the horizontal line A-A of the framework on page 7 (Lesson I), in place of the letters H, I, J, K, L, M, and N, the figures repre- senting the years in our statistical table. This gives the following horizontal scale line: A 1913 1914 1915 1916 1917 1913 I9l=i 1 A (H) (1) (J) CK) (L) (M) CM) detp:rmixing the vertical scale With the years representing the horizontal scale, the average price of bacon figures — the dci)endent variable — must necessarily be meas- ured by the vertical scale. The units of this 16 Chartography in Ten Lessons scale are determined arbitrarily by figures that must have a spread sufficient to include the lowest as well as the highest price of bacon that is to be recorded according to the statistical table. The lowest price is 26.4 cents in the year 1915 and the highest is 57.2 cents for the year 1919. It has already been stated that the lower or base horizontal line is zero and should be indi- cated by a cipher. This also means, inasmuch as the vertical lines rest upon the base line, that the beginning or start of the vertical lines must be at zero. The framework above the base or zero line G-G on page 7 (Lesson I) provides six squares within which the highest number — 57.2 — of our statistical table has to be recorded. With these facts to consider it is a simple mathe- matical computation which shov/s that the small- est unit that can be made for the vertical scale is that of 10 for each square. Placing this unit from to 60 on the vertical scale line H-H of the framework shown on page 7 (Lesson I) instead of the letters G, F, E, D, C, B, and A, gives the vertical scale on opposite page. THE SCALE LINES We have completed both the horizontal and vertical scales as determined by the figures of our statistical table. Substituting these scales on The Scales 17 the fratne-work of our chart in place of the letters designating the lines gives the results shown on the next page. ~H These scale lines — the horizontal and ^^' ^' vertical — are very important features of a chart ; in fact, without them a chart is unintelligible. They must be adapted to the arbitrary limitations of space, and this adaptation is readily brought _ (Q) about by increasing or decreasing the space allotted to each unit of each scale to correspond to the requirements of - (D) the particular statistical problem. The vertical scale unit itself can also be in- creased or decreased as the particular ~ problem requires. This scale measures, by equal distance along all the vertical _ ,^ lines, the units of the variables that are being charted — it represents Vjy space on the lines of the chart the equivalent — (6) of an agreed uj)on element of the statis- tics as determined by the units selected. READING THE SCALES The horizontal scale should read from left to right with the earliest year to be recorded appear- ing first and the remaining years following con- secutively in i>oint of tijne. 40 30 20 10 18 Chartography in Ten Lessons The vertical scale beginning at zero should read upward from the bottom or base line to the top or horizontal scale line. 1913 1914 1915 1916 1917 1918 1919 60i 50 AO 30 20 10 '■' ■ 572 49.5 36.2 26 7 26 7 264 28 1 This arrangement "faces" the chart to the left. A chart that faces to the right, faces in the wrong direction, or, putting it another way, a The Scales 19 chart that does not face to the left does not face in the "right" direction. PLOTTING THE STATISTICS We are now prepared to begin the plotting of the statistics. With the vertical and horizontal lines drawn the proper distance apart and with the figures of the years and vertical scale units cor- rectl}'^ indicated by lead pencil marks, the student next begins to plot on the respective vertical lines, by means of pencil dots, the exact positions of the figures of the statistical table as determined by the vertical scale. This scale applies similarly to measiu-ements on all the other vertical lines as much as it does on the vertical scale line itself. That is, any unit of the vertical scale line, say 30 of our present scale, has exactly the same relative position on all the vertical lines as it has on the vertical scale line. The first figure of om- statistical table that is to be located on the chart is 2(5.7, representing in cents the average jirice of a |)ouiid of bacon on April 15, 1913. The first vertical line, wliicli is our vertical scale line, also rei)resents that year, as indicated by the figures 1913 at the top of the line. Starting at the ba.se of this line at we proceed upward to 10, to 20, and .somewhere be- 20 Chartography in Ten Lessons tween this unit designation and the next one, 30, must be the proper location for the figures 26.7, THE USE OF GRADICULES It is easy to locate where 25 should be — midway between 20 and 30 — even without the aid of the slight projections or gradicules which have been inserted on the left of the vertical scale line in the chart on page 18 for the purpose of aiding the beginner. Each of these gradicules represents one-tenth of the vertical scale unit, or 1, and there are ten gradicules betAveen each unit of 10. They perform a function similar to the sub- divisions of the inch unit on the ordinary ruler — thty enable the student to locate with facility on the framework any figure of the statistical table that falls within the round numbers of the vertical scale units. With the assistance of these gradicules it is a simple matter to determine the correct location on the vertical scale line of the figures 26.7. This is indicated by means of a pencil dot. The same procedure is followed in locating on their respective vertical lines, as indicated by the vertical scale, the remaining figiu"es for each of the other six years of the horizontal scale. The locating of each number on each vertical line should be done by starting at the base or The Scales 21 zero line and counting upward, and not by start- ing from the position of the preceding pencil dot. One reason for this is to prevent the possibility of error in the location of the num- bers in case a mistake happens to be made in placing the first one on the vertical scale line. Besides, it is important that the beginner should have impressed upon his mind at the outset that all positions of numbers charted by means of a curve are determined in relation to the base or zero line. This is clearly indicated on page 18. On this drawing the numbers represented by the pencil dots, and which are those of our statistical table, are i)laced opposite their respective dots to emphasize their location. This presentation has prepared the student for the actual drawing of the curve. This he does by starting his pencil at the dot on the vertical scale line representing the number 26.7 for the year 1913, and by means of a straight line marks the space between this dot and the dot rei)rescnting 26.7 on the second vertical line, which latter, according to the horizontal .scale, represents the year 1914. It .so liai)i)ens that the average price of bacon on April 15, 1914, is identical with the i>rice on April lH, 1013, ac- cording to onr statistical tabic. This gives a straight line connecting vertical lines 191.'5 and 22 Chartography in Ten Lessons 1914 at the point 26.7. The dot on the vertical line representing the year 1915 is at 26.4, this figure being the average price of bacon on April 15 of that year. The student connects the dot . 1913 60 50 1914 1915 1916 1917 1918 1919 40 30 20 10 / / / / / YEAR CENTS 1913 26.7 1914 26.7 1915 26.4 1916 28. 1 1917 38 2 1916 495 1919 572 --^ / representing 26.7 for the year 1914 with the dot at 26.4 for 1915. Continuing this process for the remaining dots gives the curve shown above. The Scales 23 In this drawing the lead pencil dots have been erased, as have also the gradicules along the vertical scale line shown in the chart on page 18, these dots and gradicules being of no further use. The figures representing the price of bacon for the different years have also been removed from their positions opposite the dots and have been placed in statistical table form, with the years in the first and the prices of bacon in the second columns. QUESTIONS FOR SELF-EXAMINATION 1. What is the essence of a chart? 2. What are variables? What is an independent varia- ble? A dependent variable? 3. What are scales? Describe the horizontal scale. The vertical scale. 4. What is the relation between the variables and the scales? 5. What scale does the independent variable take? The dependent variable? 6. How is the vertical scale determined? 7. What is the base line? What service does it per- form? What is the zero line? What is the relation between the lower horizontal line and the zero line? Between the lower horizontal line and the base line? 8. What is a square? How is it formed? What is its function in chartography? 9. What is the vertical scale unit? How is it deter- mined? 10. What is a vertical scale line? A horizontal scale line? What relation to these are the vertical and hori- zontal scale units? 11. How should the horizontal scale be read? The ver- tical scale? In what direction should a curve chart face? 12. What is meant by plotting the statistics? How is it done? 13. What is the relation of the vertical scale units to vertical lines other than the vertical scale line? 14. What are gradicules? Of what use are they in plot- ting the statistics? Where are they located? Of what use are pencil dots in plotting the statistics? 15. How are the positions of the figures of the statistical table on the framework determined? What service does the zero line perform in this determination? 24 LESSON III The Curve Chart Making the Curve Heavier — The Horizontal Scale Unit — Squares Should be Equal — Effects of Different Scale Units — Dropping the Zero Line — Indicating the Absence of the Zero Line — Divisors for the Vertical Scale. In drawing the curve remember to make it heavier than any other Hne. The purpose of this is to have it stand out prominently and so catch and hold the eye of the reader. The curve should he the most conspicuous of any line on the chart for the reason that it embodies or symVjolizes the most important facts that are presented — it is the why and the wherefore of the chart being called into existence. Con- versely, the framework lines making up the back- ground of the chart, that is, the horizontal and vertical lines, should be drawn with a lighter touch of the pencil to paj)cr. The curve is a continuous, unbroken line, and has its origin at the i)oint along the vertical scale line that is determined for the time or other designation of that line by the statistics and the vertical scale. It moves across the page from point to point on the vertical lines and in the 25 26 Chartography in Ten Lessons direction from left to right as the respective numbers of the statistical table determine. The curve terminates on the last vertical line at the point the statistics require. It takes the shortest distance between two points and generally should approach each slantingly. THE horizontal SCALE UNIT In a curve chart the unit of the horizontal scale element — in our present case this is a calendar year — marks a point as distinct from space between points. Each vertical line pro- jects or extends its horizontal scale unit down- ward all along the entire distance of that line even to the base or zero line. The curve cannot and does not affect it — the curve does not move any horizontal scale unit a hair's breadth from its place on a particular vertical line. Or, rather, the horizontal scale unit does not follow the curve from its point of contact with it on one vertical line to the point of contact with another horizontal scale unit on another vertical line. For instance, the year 1913 ends with the vertical line so designated and does not cover the space between vertical lines 1913 and 1914. Quite commonly in curve charts this distinction is overlooked, particularly by beginners, and the horizontal scale element is sometimes made to The Curve Chart 27 represent space on the horizontal scale line and between the vertical lines. This is a mistake. SQUARES SHOULD BE EQUAL In a curve chart it is desirable to have the curve move from point to point in squares or areas of equal spacing in all directions, whether these be large or small. This means that both the horizontal and vertical scales should be deter- mined upon a basis that will permit equal spacing between the units of each scale, that is, between the horizontal lines of one scale and the vertical lines of the other. This allows the curve to move up or down and from left to right an equal distance for each unit of measurement of both .scales. Many curve charts are being made in which this rule is violated. It must be added, however, that the observance of this principle is not always possible owing to the arbitrary limitations of .space and to the necessities of the scales. The problem for the chartographer is to secure as accu- rate an observance of this rule as his difficulties will permit. He should constantly koej) in mind the important fac-t that tlie horizontal and vertical lines are made use of to measure the quantity or volume or other specified quality of the .statistical element that is charted, and that these rules of measurement should be as fair as possible. 28 Chartogkaphy in Ten Lessons It is recommended that the beginner at first draw liis frame work or scaffolding lines, both hori- zontal and vertical, exactly one inch apart, thus giving square inches within which the curve moves. Each scale will then have its units of measurement one inch distant from each other. Later on the student can practice with lessening or lengthening this distance, keeping in mind not to move the scale units to points less than one- half inch or further apart than one and one-half inches. He should not permit the units of either scale to be separated by any greater distance than the units of the other scale. EFFECTS OF DIFFERENT SCALE UNITS The student should also practice changing the unit of the vertical scale within the inch square, increasing or decreasing it to other selected units, in order to observe carefully the effects these dif- erent units have upon the movement of the curve. In the drawing on page 22 the unit is 10. Let us substitute for it the unit 5, as in the drawing on the opposite page. A study of these two drawings will disclose a number of important differences. Probably the most important of these is the fact that a vertical scale unit one-half as large, other factors remaining the same, doubles the space within which the curve moves. Conversely, The Curve Chart 29 doubling the scale unit decreases by one-half the distance the curve moves. This space in the drawing on page 22 requires 1913 1914 1915 1916 I9l7 1918 19 60 55 50 45 40 35 30 / / / 1 / ( / ' 9«; vertically a fraction more than tiircc of the squares — the Hj)rea(i in the difrcronce between the lowest and highest numbers of the statistical 30 Chartography in Ten Lessons table is 30.5 and with 10 as the unit this leaves .5 more than three times 10. In the drawing on the preceding page the vertical scale unit 5 requires a fraction of .5 more than six times 5 to accommodate the curve, or seven vertical squares at the very least. As the drawing on page 22 provides only six squares, another one has to be added to the framework, as is done on the page preceding. This is accomplished by inserting an additional horizontal line, either above the top or below the bottom horizontal line, and then extend- . ing to it all the vertical lines. With the new vertical scale unit being 5 and with the highest number to be charted being 57.2 for the year 1919, a square must be provided for each of the units of 5 if the scale is to begin at zero. This demands at least twelve squares for the vertical scale from to 60. But it is physically impossible to accommodate this many squares of the present size within the space limitations. DROPPING THE ZERO LINE The next step is to ascertain from the statis- tical table the lowest number to be recorded. This is 26.4 for 1915. It is clear from this that the space occupied by all the squares below the vertical scale unit 25 will not be needed for record- ing the movement of any part of the curve, for The Curve Chart 31 in not a single year of all the seven given in the statistical table does the price of bacon fall below that unit. Consequently, beginning the vertical scale at 25 instead of at permits the elimination from the framework of iSve squares. The number that remains, which is seven, is sufficient for the requirements. It has been made clear in preceding lessons that the bottom horizontal or base line of a curve chart represents zero of the vertical scale and is indicated by a cipher as follows: INDICATING THE ABSENCE OF THE ZERO LINE Such a line, of course, cannot possibly be used as the base line with the unit of our vertical scale starting at 25, so the zero designation is omitted. Attention should always be called to this omis- sion on the chart itself and this can be done by inserting directly below the base lirje, witii its proper unit designation, a faijit lino of dashes or one of dots, or a wavy or slightly undulating line, as indicated on the next page. Rulers pro- vided with these undulations can be purchiised. 32 Chartography in Ten Lessons The student should keep in mind as a general principle the fact that the vertical scale begins on the base line at 0, although he will frequently find that this is physically impossible because of the nature of his statistical problem. This pre- vails more often among large numbers than with percentages. Usually the lowest nu nber to be charted starts at a point so high above that the space required to show the latter on the chart is out of all proportion to that necessary to in- dicate the movement of the curve. Again, frequently in such cases the vertical scale unit determined by including zero becomes so large that fluctuations in the movement of the curve reflecting the trend of the statistics (which fluc- tuations would be made clear by the use of a smaller unit) are smoothed out or flattened so that that which should be a curve approaches nearer to a straight line. Thus it is not always possible to plot a curve chart so that the zero of The Curve Chart 33 the vertical scale will be shown and at the same time clearly present the trend of the statistics, which latter is the primary object of the curve. In beginning to read a curve chart, among the first things to be observed is whether the vertical scale starts at zero and if it does not to make proper allowance for this fact in the inter- pretation of the movement of the curve. Unless this is kept in mind an erroneous idea or impres- sion of the extent of the movement will result. A chart that does not present the zero line and fails to call attention to the omission in the ways indicated, or neglects similar precautions, is constructed in error. Such a chart is very likely to be misleading no matter how excellent or perfect its other features may be. DIVI.SORS FOR THE VERTICAL SCALE The selection of the vertical scale unit is thus not without its difliculties. These the student will learn to overcome as his experience with varying stiitistical problems increases. lie will learn, among other things, that particular numeri- cal divisors are more advantageous as units than some of the others. The divisf)r 3, for instance, is an awkward and inconvenient scale unit, not only for computing on the vertical lines the measurements of the 34 Chartography in Ten Lessons statistical element but also for calculating by the interpreter of the chart. The divisor 2 is much better, and 5 and 10 are nearly always ideal. Such units as 3, 4, 6, 7, 8, and 9 are not as good as 2, 5, 10, 20 and so on, the latter group being more easily divisible into the spaces along the vertical lines as well as into the numbers of the statistical table. Whatever scale unit is selected it must permit the inclusion within the arbitrary limitations of the framework of the smallest as well as the largest number that is to be charted. The unit must be such as to permit of a spread between the lowest and highest numbers charted sufficient to bring out clearly in the curve the points or tend- ency to show which the particular chart has been designed; at the same time it must not be too small as to result in exaggeration. It is as serious an offense to exaggerate with curves as it is with words. Accuracy in chart expression is as im- portant as is the use of words in expressing thought, and the various uses or functions of the vertical scale unit have much to do with accuracy in curve charts. On a finished chart the student will not find any dots and similar marks used as guides in erecting the scaffolding of the framework, which means that all such marks must be erased from The Curve Chart 35 the completed chart. He will find, however, that all the vertical and horizontal lines make complete right angles at all points of junction; that all such lines are straight lines; that they form accurate squares; that the curve is slightly heavier than the other lines; that the scale unit figures are in their correct positions in relation to their respective lines; and that the horizontal and vertical scale unit figures do not crowd the lines but are separated from them by the correct spacing. QUESTIONS FOR SELF-EXAMINATION 1. Why is the curve made heavier than other lines? 2. Define a curve. How is it emphasized in comparison with the horizontal and vertical lines? 3. Define the horizontal scale unit. What function does the vertical scale lines perform for this unit? What is the relation of the curve to it? 4. What effect have unequal squares on the movement of the curve? What relation is there between the squares and the scale units? 5. What are some of the effects of changing the vertical scale unit? 6. When is the zero line omitted? How is this omission indicated? 7. Explain the reasons for omitting the zero line. 8. What effect has the omission of the zero line on the reading of the curve? 9. What are numerical divisors? When and how are they used? What ones are better than others? 10 What must the divisors provide for? 36 LESSON IV Features of a Complete Chart The Statistical Table — Table Should Appear on Chart — The Make-up of the Table — Spac- ing the Columns — The Form of the Table — Duplicating the Scale Units — The Place for the Horizontal Scale — Ward Designation of the Scale— The Title— The Foot-Notes—The Neat Lines. The drawing on page 38 is a complete curve chart constructed according to the instructions of the preceding Lessons. The student should examine carefully every one of its features. Particular study should be given by the student to the statistical table. It occupies the position in the lower right hand corner of the drawing on page 22 (Lesson II) but in the accompanying chart it is located in the upper left hand corner. In each ca.se the location of the table is adapted to the requirements of the fjarticular chart and each is correctly i)laced. It will be found that one or the other of tlie.se two positions is usually the place for the table, the lower left hand corner and the upper right hand corner nearly always being required for the free and unobstructed movement of the curve. S7 (J 38 Chartography in Ten Lessons THE AVERAGE PRICE* OP BACON UNITED STATES. I9I3-I9I9 1913 60 55 50 45 40 35 30 25 1913 1914 1915 CENTS 1916 1917 1916 / YEAR CENTS 1913 26 7 1914 26 7 1915 264 1916 28.1 1917 382 1918 495 1919 57.2 / 1 / / / ^ 1914 1915 1916 1917 1918 1919 60 55 50 45 40 35 30 25 1919 d^arkstics art rr&m Menrtily U&bo*- R«vl«<M, p, 77. I • Av«rAg« prtc* itt dt Apf tl IS oF *ach ywar. This simply means that there is no arbitrary position on the chart for the statistical table but that its location is determined by the result of the plotting of the curve. The only general rule to follow is to place the table of figures in the Features of a Complete Chart 39 particular position on the chart that disjtlays it to the best advantage without at the same time crowding the scale lines or interfering with the curve. Breaking the vertical lines 1914 and 1915 and the horizontal lines 45, 50, and 55, as is done in the accompanying chart, is not objection- able but rather advisable in preference to these lines extending through and breaking up the table. "Boxing" the two columns of the table with light lines, as in the accompanying chart, adds to the neat appearance of the finished diagram. TABLE SHOULD APPEAR ON CHART Virtually every chart is based upon statistical information. Usually this infortnation is in the form of a statistical table or colunms of figures. If the chart has been properly constructed and if the figures of the table are correct, the presence of the statistics is not e.ssential to a complete understanding of the chart — its meaning Mill he clear without the figures. Nevertheless it is highly important in good chart making that the statistics uf)on which the diagram is based should occupy an inifjortant i)lace on the chart. This reproduction of the fignrcs furnishes proof, if proof is needed, of the correctness of the move- ment of the curve as shown in the chart and will al.so be of service to those who nuiy wish to u.se the 40 Chartography in Ten Lessons data in other directions or to make different com- pilations. Unless the statistics from which the chart is made are shown upon it there is no easy way to check up the work of the chartographer. THE MAKE-UP OF THE TABLE The internal make-up and arrangement of the statistical table also require some thought from the student. Its construction would appear at first glance to be an easy thing to do and yet the task has its difficulties. All the numbers of the same statistical element that are to be compared should be placed in the same vertical column one under the other and not too far apart, the digits of the tens and hundreds and so on occupying their proper positions in relation to similar digits of other numbers in the same column. In the case of years, these are ar- ranged vertically and in proper sequence of time one under the other with the earliest year at the top and the latest year at the bottom of the col- umn. Nearl}^ always the years are in the first column to the left in a table of two or more col- umns. Vertical columns of figures read downward from the top and never upward from the bottom. This is in inverse order, it will be noted, to the eading of the vertical scale units. Each column has its proper word designation Features of a Complete Chart 41 just above the first number, as years and cents in the table of the chart on page 38. This is the column heading. The space for it is usually very limited and for this reason it should be confined to simple words of the fewest possible letters con- sistent with clearness as to the meaning of the column of figures. Double meaning of words should be as carefully guarded against as indefin- iteness in meaning, each being a serious offense against clearness of expression. \Mien two or more words are necessary in the heading of a column it is usually advisable to make of them two or more lines just above the first number, with each word having a line to itself instead of all the words occupying a single line, which latter nearly always extends the heading too far on either side of the column of figures. SPACING THE COLUMNS Where two or more columns of figures are in the same table attention has to be given to the proj)er si)acing between the columns as well as between the numbers themselves and their headings. Hut in every table the number of columns sliould be strictly limited to the fewest possible for the purpose in view, the inclusion of any that are not necessary detracting froni the em[)hasis that must be given to the principal 42 Chartography in Ten Lessons facts and tendencies shown by the statistics. The table must be complete in itself, however, with no vitally important factor missing. To this end more than one comparison should not be at- tempted in the same table. The form of the table has to be adjusted not only to the size of the chart but in particular to the space available on the framework for its presentation without interfering with the curve. Interference by the table with the light horizontal and vertical lines is not so important; nor is a correct interpretation or reading of the curve interfered with even when the vertical scale line is broken into by the table at points which the curve does not approach. THE FORM OF THE TABLE There are distinct forms best adapted to par- ticular purposes with which the student will be- come familiar only by practice. He will have to decide at times whether he will include all his data in one table or break them up into two or more tables with a chart to illustrate each. Com- pactness as well as proximity of the numbers for comparative purposes are advantages which must sometimes be surrendered at the demand of more pressing requirements. If the table is too large confusion to the eye results and difficulty is Features of a Complete Chart 43 encountered in following the significance of the separate columns. Interpretation also is particu- larly taxing if the tabulation is dealing with a complex mass of figures. SIMPLICITY the guiding RULE If the student will keep in mind that simplicity must be the guiding rule he will usually not go far WTong in his decisions. Much, of course, depends upon the data that must be shown but quite often more can be eliminated from the columns than at first seems possible. Foot-notes as explanations of the table often show a way out but these should be kept at the minimum. A title to the table separate from that of the chart is imnecessary. Sometimes the exigencies of space limitations combine witli the requirements of the movement of the curve to prevent the use of the form of statistical table that has been described and which is shown in the chart on page 38. In such cases recour.se has to be had to some other form, some- times to that shown on page 4 (Lesson I) with the years and j)rices of bacon arranged in hori- zontal lines instead of vertical columns. At times when this form has to be resorted to it is not possible to f)Iace the table within the frame- work and in .such cases a position has to be found for it elsewhere on the chart. 44 Chartography in Ten Lessons duplicating the scale units In the chart on page 38 the unit figures of the horizontal scale have been placed on both the top and bottom lines and those of the vertical scale on both the first and last vertical lines. This arrangement has many advantages. While it is not essential to the reading of the curve, at the same time it facilitates the interpretation of its movement in that the reading of the prices of bacon for the first and immediately succeeding years is permitted without requiring the eyes to move upward to observe these horizontal scale units; also, it permits the quick reading of the prices of bacon for 1919 and the immediately preceding years without requiring the eyes to travel across the page to observe these particular vertical scale units. The scale units arranged along all four sides also serve to give a border- like appearance to the framework and introduce a little greater uniformity in place of a tendency towards a lack of balance. THE PLACE FOR THE HORIZONTAL SCALE Some chartographers prefer placing the hori- zontal scale units along the bottom line only. My own preference is that where they are to appear only once on the chart then the place Features of a Complete Chart 45 for them is on the top line. That line will be found to be much more convenient as the hori- zontal scale line; besides, there are other im- portant uses for the bottom horizontal line, such as serving as the base line and as the zero line. This preference is influenced also as the result of more than ten years' experience in chart making for practical commercial purposes. Dur- ing this experience it has been observed that most people in reading a chart start at the top with the title and glance downward. With the units of the horizontal scale on tlie top line the reader early in the process of interpretation is informed of these important facts which he must know if he is to read the chart intelligently and correctly. Placing the horizontal scale units on the bottom line meets with the objection that the space beneath this line is usually needed in most curve charts for important exj)lanations and foot- notes, such as credit for the source of the statis- tical information upon which the chart is built, notice of coj)yright, and so on. With the hori- zontal scale figures also there that section of the chart is likely to give the a})pearance of crowding. Again, with the horizontal scale figures located on the bottom line I have frequently encountered practical difficulties hard to overcome because the first and last of these units interfere with 46 Chartography in Ten Lessons those of the lowest scale measurement of the vertical lines, both sets of figures being located at nearly the same point of the right angles formed by these vertical and horizontal lines. As op- posed to this, it is nearly always possible to extend the upper part of the framework at least one series of squares beyond the highest point to be recorded by the vertical scale and this permits the figures of the horizontal scale units to have a line all to themselves without interfering with and without interference from any of the figures of the vertical scale. There is no objection, of course, to reproducing the horizontal scale figures on the bottom line whenever there is room for them, and this prac- tice is recommended as being advantageous, especially in charts of unusual depth, as it facili- tates a quick reading of the curve movement. WORD DESIGNATION OF THE SCALE Further assistance in the interpretation of the chart, and especially in the reading of the curve, is rendered if the primary characteristic of the statistical element represented by the vertical scale is indicated by a word designation just inside the top border line and directly above the center of the horizontal scale line. This is shown in the designation "Cents" in the chart on page Features of a Complete Chart 47 38. Such a designation states concisely to the reader what the vertical scale figures represent — it explains the essence of the curve. Its value and usefulness will be impressed upon the student as he progresses in his studies. THE title Another important matter to be considered before our chart is a complete one is the title or heading. Every chart must have a title. With- out it the chart is almost as incomplete as it would be if the curve itself were omitted. The title is as much a part of the chart as are the scale lines or table of statistics. It is more im- portant than the beginner in chart making is apt to realize. The title should not contain a single unneces- sary word. The space for it is usually limited and too many words detract from the effect in expressing the idea intended. Simple words of one syllable are preferable. This choosing of words in the selection of a title is a splendid exercise in enabling one to secure a better com- mand of the English language and in compre- hending more clearly himself the essence of the chart. The title should be so clear in its meaning that misinterj)retation is impossil)lc, and so com- prehensive in its scope as to cover all the import- 48 Chartography in Ten Lessons ant data presented by the chart so that the inter- preter will not have to look elsewhere for explana- tion. This is not alwaj's possible and in such cases a foot-note explanation at the bottom of the chart is advisable, Indefiniteness in title mean- ing is a serious offense. Similar statements are equally applicable to any sub-title. The position of the title is, of course, at the top or head of the chart, as shown on page 38. The best title is one comprising a single line but this is not always easy to accomplish. In the chart referred to it has been found necessary to have two lines in the title, and in such cases the letters of the words in the second line should be slightly smaller than those of the first line. The principal idea in this chart is the tendency in the price of bacon, so its title becomes "The Average Price of Bacon." But as this does not give the information quite complete enough, the reader is told in the second line that the price is for the entire "United States" and for the years "1913 to 1919." The asterisk after the word "price" refers the reader to the foot-notes, where it is stated that the average price given is of April 15 of each year. THE FOOT-NOTES The place on the chart for the foot-notes is Features of a Complete Chart 49 just below the base line and outside the frame- work proper. These serve a useful purpose in presenting descriptive information sometimes necessary to clear up a point that has not been brought out sufficiently in the chart, as indicated in the use of the asterisk in the chart on page 38. In the foot-notes there should always be a state- ment as to the source of or authority for the statistics upon which the chart is based. This is shown on the chart just referred to by the notation "Statistics are from Monthly Labor Review, p. 77, U. S. Bureau of Labor Statistics." The foot-notes also supply a convenient place for the legal statement required in case the chart is copyrighted. THE NEAT LINES With the drawing of the border or "neat" lines, one on each side of the framework and usually about one-half an inch from the horizontal and vertical scale lines, the chart is comi)leted. These neat lines give a sort of frame to the chart, as shown on page 38. In a good curve chart the principal conclusions to be drawn from the statistical table are made plain, all doubt as to the tendency or course of the phenomena rei)resented by the numbers is re- moved, and all i)Ossible errors have been elimi- 50 Chartography in Ten Lessons nated. This ability to analyze the significance of a table of statistics, to interpret the results correctly and clearly, and to indicate the con- clusions lucidly and succinctly is one of the characteristics of chartography. The results disclosed by a curve based upon a statistical table quite often reveal at a glance important facts that could not have been known except from considerable study of the figures by an expert. Usually an accompanying explanation or analysis is unnecessary. If so the chart has failed of its primary purpose. The task of checking-up is not optional with the student; it is compulsory — he not only should but he must go over carefully each chart from top to bottom. If it is a curve chart, do all the horizontal scale units center on the end of the vertical lines.'* Are the respective vertical scale units on the right in the curve chart directly opposite and at the end of the same horizontal line as those on the left.? If the curve chart contains a zero or 100 per cent line, has it been made wider or heavier than the other horizontal lines? If the zero line is not shown, does the bottom or base line clearly indicate that the vertical scale does not begin with zero? Features of a Complete Chart 51 If it will aid in the easy reading of the curve see that the horizontal scale figures are duplicated at the base line. In most curve charts it is best to have the vertical scale figures on the right as well as on the left vertical scale line. If only one set of scale figures are used, however, these should be alongside the first or left vertical line. Do not make use of the first and last vertical lines of a ciu-ve chart as the neat lines of the frame. They should be reserved strictly for the vertical scale units and should be no heavier or wider than the interior vertical lines. Do not forget that it is the independent variable that takes the horizontal scale, especially in curve charts involving periods of time. Follow each curve from its beginning on the left to its termination on the right to see that it is continuously correct — that there is no "break" in it. See to it also that the curve is a slightly heavier line than the vertical and horizontal lines. 1 T QUESTIONS FOR SELF-EXAMINATION 1. Describe in general terms the most important char- acteristics of a curve chart. 2. What is the statistical table? What is its relation to the curve? 3. What is the position of the table on the chart? What are its general features? What is meant by "boxing?" 4. What is the internal make-up of a table? What is a column heading? What is meant by spacing? 5. What is the guiding rule in table construction? C. What features of the chart affect the form of the table and its position on the framework? 7. What is meant by duplicating the scale units? How is this done? What are the advantages? 8. What is the position for the horizontal scale? Give reasons supporting your statement. 9. What is the function of the word designation of the vertical scale? 10. What Is the title? What is its location? Describe the principles underlying the selection of words for the title. 11. What is an asterisk? What are its uses in chartog- ' raphy? 12. What service do foot-notes perform? Where are they located on the chart? What do they usually comprise? 13. What are neat lines? What is their position on a chart? 52 LESSON V The Bar Chart Making Bars from a Curve — Making a Curve from Bars — Advantages of the Horizontal Bar — Reversing the Scales — Width of the Bar — Separation of the Bars — Location of the Table — The Bar and the Curve. Emphasis thus far has been placed on the curve as the method of expression offered by the art of chartography. But tliere is also the bar. Many of the principles of construction already explained in describing tlie curve apply with equal force to the bar chart. In fact, there are many points of similarity between these two different kinds of charts. MAKING BARS FROM A CURVE This tlie student will be able to realize clearly by taking a curve chart and drawing vertical bars from its base line to the points of the curve. This has been done in the chart on page 54. It is merely the result of taking the curve chart on page 38 (l/csson IV) and with as few changes as possible transforming it into a bar chart. In order to secure a bar for each of the seven years it is necessary to add another vertical line 5S 54 ClIARTOGRAPHY IN TeN LeSSONS to the right of the one for the year 1919 and extend to it the top and bottom horizontal lines. The horizontal scale unit for each year is moved to the right from its former position at the top of a vertical line so as to occupy space between the vertical lines and to be above the top of the bar. No change is made either in the unit of measurement of the vertical scale or in its location. The Bar Chart 55 making a curve from bars This procedure enables the many points of similarity between the curve and the bar chart to be quickly recognized. This similarity will all the more be indelibly impressed upon the mind of the student if he will take the vertical bar chart on page 54 and diaw a continuous curve from left to right touching the tops of all the bars. Then if he will cut out a piece of blank paper so that its upper edge conforms roughly to the curve he has drawn he will find, by placing this on the bars, that the latter are hidden from view and that the curve remaining in sight ex- presses just as clearly the tendency shown by the bars. In other words, his cut piece of blank paper has simply restored the original curve chart on page 38 (Lesson IV). This procedure also emphasizes strikingly the essential difference between these two kinds of charts. Tliis difference lies primarily in the fact that the horizontal scale of a curve registers points on lines while the horizontal scale of a vertical bar chart registers space beiireen j)oints on lines. But in changing the curve to the bar we have not secured a good bar chart. In the first place the bars are entirely too wide to represent such .small amounts as cents. In the .second place the bars take up entirely too much space — the same 56 Chartography in Ten Lessons ends can be accomplished by the use of a narrower bar. In the third place the result is a vertical bar, that is, a bar standing upright on its end, A horizontal bar, that is one lying on its side and extending from left to right, is preferable. ADVANTAGES OF THE HORIZONTAL BAR This preference is based on an experience of years in meeting the every-day problems of chartography. It convinces the writer of the greater utility of the horizontal bar. Quite probably there are occasions when it is advisable to have recourse to the vertical bar, but at the same time where there is a choice between the two the horizontal bar will be found to be more advantageous. It gives greater opportunity for the display of letters and figures where the limita- tions of space or other considerations require that these be placed on the bars themselves. In such instances, in order easily to read the words or figures on vertical bars the chart usually has to be turned half way round to the right, whereas if the bars are horizontal the figures and letters read in the natural direction. In brief, with the vertical bar the chartographer will encounter more difficulties than with the horizontal bar in the placing of his table, figures, and letters. The ability to select advisedly in those cases where it The^Bar'Xhart 57 might be advantageous to employ the vertical bar will come to the student with practice and experience. It is recommended that in the mean- time he confine himself to the practice of the horizontal bar. Such a bar chart is presented on the foUow^ing page. It will be observed, from a comparison of its statistical table with that of the curve chart on page 38 (Lesson IV), that it is constructed from the same set of figures. REVERSING THE SCALES The horizontal bar has necessitated a reversal in the location of the scales in comparison with those of the curve. Instead of the independent variable — the years — occupying the horizontal scale position it takes that of the vertical scale, and the dependent variable — the prices of bacon — becomes in turn the horizontal scale. This permits of the measurement of the movement of the bars from left to right and not from the bottom up, as with the vertical bar. Otherwise we could not secure the advantages of the horizontal bar. In a horizontal bar chart the figures of the vertical scale, quite froquontly comprising periods of time, are located directly to the left of the beginning of the bars, the figures for each year being centered adjacent to their respective bar. 58 Chartography in Ten Lessons A "T in « r. OD (J> Ok 0) o> 0) O) 01 0> • >• 3" t 3 o I The Bar Chart 59 The last digit of the number should not be per- mitted to crowd the end of the bar too closely. WIDTH OF THE BAR It will be noted that the bars are narrower in the chart on page 58 than in the one on page 54. This feature of the bar is important. Just how narrow or how wide or deep the bar should be will depend upon a number of factors, such as the nature of the particular statistical problem, the arbitrary limitations of sj)ace, and so on. No definite rule can be given except that all the bars of a chart must be of uniform width, they should be sufficiently wide to be easily seen, and they should convey an impression of the volume or quantity represented. P'or instance, a bar representing billions should be wider than one representing millions; the latter wider than one representing hundreds of thousands; and the latter of greater width than one rc})resenting thousands, and so on. Wide bars are preferable to too narrow ones. In l)egirming to draw the bars the sttident should i;i<licate on the sheet by light lead pencil lines instead of dots the width and length of each bar, the former being arbitrarily determined by the number of bars that is to go in the available space and the latter by the quantity or volume each bar rei)resents as determined by the statis- 60 Chartography in Ten Lessons tics and the scale unit. It is first advisable to determine from the statistics and the horizontal scale unit the length of the shortest and the longest bar. All the other bars fall within the limits these two set. Begin plotting the chart with the bar for the earliest year at the top and just beneath the horizontal scale line, outlining the bars downward as the years determine. These skeleton bars should then be filled in black by rotating the pencil point within the outlines. separation of the bars Between each bar representing the statistical element of the vertical scale there should be a separation sufficient to distinguish it from the preceding and following bar. In the chart on page 58 this has been done by leaving in the original drawing a space equal to one-tenth of an inch. Usually this is too much spacing. Besides, it requires a greater amount of painstaking labor than should ordinarily be given to a bar chart. How this labor can be eliminated the student will be informed in a succeeding Lesson. The bar chart on page 58 shows the vertical lines extending from the points of the scale units on the top horizontal line to the base hori- zontal line except where the bars and the statis- tical table intervene. These extend downward The Bar Chart 61 the units of measurement of the horizontal scale to each of the seven bars at the various points of contact of the vertical lines with the bars. Ordi- narily these vertical lines should not be extended between the bars but to the first bar only that interferes with their further extension. These lines are permitted to be seen on this chart merely to inform the student as to the purpose of the vertical lines in a bar chart. All sections of verti- cal lines tliat have been drawn within the bars should be pencilled out of observation as the body of the bar is pencilled in. LOCATION OF THE TABLE The location of the statistical tal)le in the upper right hand corner is that which will usually be found best adapted for this use. This is true be- cause this position, as a general thing, is opposite the shortest bars and thus has the largest area of unoccupied space. The lower right hand corner is frefpiently taken up with the extension of the longest bars rei)resenting tlie largest numbers to be charted, and the upi)er and lower left hand corners always contain the })eginning of the bars. In cases where the extensions of the bars from the earliest to the latest years show a decrease instead of an increase, the statistical table s}K)nl(l be located in the lower right hand corner. 'J'he 62 ClIARTOGRAPHY IN TeN LeSSONS table should be "boxed," that is, enclosed in a light frame composed of two vertical and two horizontal lines connecting at their ends and form- ing right angles. Separating the nunibers from their table and placing them on the individual bars adjacent to the figures of the years will sometimes be found advan- tageous. THE BAR AND THE CURVE As between the bar and the curve chart the latter will be found to be much more useful as well as more adaptable to a larger number of statis- tical tables or problems. It is true that in many instances either may be employed with equally successful results. The bar chart, however, is the most common at the present time not only because it is the simplest to construct but also to interpret. Its advantage lies in its simplicity — the amount or quantity or statistical element is simply represented by the length of the bar. This gives only one dimension to be read and in conse- quence there is little ground for misinterpreta- tion. As a general statement the bar method should be used where the numbers represent large volumes or quantities. At the same time there are special problems in chartography which the curve chart alone will The Bar Chart 63 solve to the best advantage. Just what are the particular characteristics of these problems the student will learn by experience. The kind of chart that will best bring out the true significance of a statistical table is the one to select. It can be said generally that with statistical tables having numbers representing very large amounts, such as billions and millions, the bar chart is preferable. Conversely, where the numbers represent small amounts, such as hundreds and tens, the curve chart is usually the best. One reason for this is that the bar conveys the idea of volume to a greater de- gree than does the curve. The difference between these two kinds of charts is strikingly presented by Brinton in his Graphic Methods for Presenting Facts. He first com- pares bars representing years or other intervals of time with j)rogress photographs. Though the bars and progress photograi)hs are valuable, he says, they give information only in spots. Then he says: "A moving-picture machine shows pictures so rapidly that the pictures blend into a continuous narrative in the eye and the brain of the observer. What the moving-[)icture is to separate progress photographs, the curve is to detached bars repre- senting time. In just .so much as the moving-pic- ture is superior to separate pictures shown by lantern slides, in just that much is a curve superior 64 Chartography in Ten Lessons to a series of horizontal or vertical bars for the same data. Unless a person knows thoroughly how to read and how to plot curves he cannot hope to understand the graphic presentation of facts." Brinton also says: "A curve permits of finer in- terpretation than any other known method of presenting figures for analysis — it gives informa- tion which many persons might not fully grasp if only a column of figures were used." And again the same author says: "One of the chief advan- tages of the curve method of presenting informa- tion is that a curve forces one to think." It will be found that plotting the curve is sim- pler than plotting the bar. It also consumes less time. Many chartographers prefer the curve to the bar method of presenting statistics because it not only brings out the fluctuations from year to year more clearly to the eye but also enables the reader to grasp more readily the tendency shown. The curve is gradually supplanting the bar in popular usage because of its greater clear- ness, and this tendency is likely to grow stronger as its advantages over the bar are more generally recognized. QUESTION FOR SELF-EXAMINATION 1. Describe the similarities and dififerences of the curve and bar chart. LESSON VI The Tools of the Chartographer Cross Section Paper — The Lead Pencil — The Kind of Ink — The Ruling Pen — Correct Posi- tion for Holding Pen — Pen Points — The Draw- ing Board — The T- Square — The Triangle — The Engineer's Scale — The Dividers — The Essential Tools. If the beginner has profited to the full extent from a careful and painstaking study of the pre- ceding Lessons he is now qualified to drop the blank sheet of ordinary paper, the lead pencil, and the common ruler and take up the real materials and tools of the chartographer. The proper use of these materials and tools will measurably facilitate and make less difficult the mechanical work of chart making and will also result in much better workmanship. It permits of the chart becoming permanently valuable as well as of its rei)roduction in any number desired. CROSS SECTION PAPER The most imi)ortant of the essential materials is the cross section or coordinate paper. A sample illustration is shown on the following 05 "" r p"* i I^V "^ "" ~" ■" "" ^ f ■- T """ i 5 i— 1 1 _ ! ( ^^ ^ Tools of the Chartographer 67 page. This section paper comprises minute squares formed by horizontal and vertical lines. On the most commonly used section paper each minute square measures one-tenth of an inch. One hundred of these squares make up a larger square of one inch, the border lines forming the square inch being slightly heavier than the other horizontal and vertical lines. Section paper can also be secured that has other rulings, such as eight minute squares each way or sixty -four to the square inch, and six each way or thirty-six to the square inch. Cross section paper thus presents a system of squares whose lines permit the easy measurement or determination, by means of space or distance on the sheet, of the quantity or volume or what- ever element it is the statistical table represents. By combining squares, space units of measure- ment as extended in both directions as the par- ticular i)roblem requires are readily determined. A sufficient quantity of section })aj)er for most charting purposes can be obtained at almost any first-class stationery store. If a large number of different charts is to be made the varying scales will likely require different subdivisions of the square inch and as it requires too nnich detail labor for the chartograi)licr himself to draw these subdivisions, it is advisable for quantity produc- 68 Chartography in Ten Lessons tion to keep on hand a supply of coordinate sheets with the diflPerent ruhngs. Even then the chartographer will not always have paper with the ruled spaces exactly corresponding to his requirement, and in such cases he will have to do the ruling himself. In sheet sizes the section paper is usually 17 by 22 inches. These sheets can be cut to meet almost any ordinary requirement; or two or more can be pasted together along the edges to meet the demand for a larger surface than that commonly required. Built-up sheets of paper can also )e formed from remnants by pasting. If a larger section-ruled surface than 17 by 22 is frequently required it will be advantageous to purchase the coordinate paper in rolls, in which form it is also prepared commercially. The section paper used should be of the best quality. There are cheap grades on the market but these do not take the ink satisfactorily and have other defects, so that in the long run it pays to purchase the better grade at little higher prices. Of course, a higher price does not neces- sarily mean a better grade, but it usually does. The section paper best adapted to ordinary chart work has the horizontal and vertical lines ruled in blue ink. On some section paper these lines are in green or purple but these colors are Tools of the Chartographer 69 not so desirable, as they are likely to reproduce lines on the photographed chart that should not be showTi. Paper with a soft surface should also be avoided as it will not take the ink properly, and from now on we are to make all our charts with pen and ink instead of pencil. THE LEAD PENCIL This does not mean that the student will have no more use for the lead pencil. In fact, he will continue to have constant need of it. The lead should )e neither too soft nor too hard — it should not be so soft as to crumble, or so brittle as to snap in two or so hard as to penetrate or puncture the drawing sheet. The best grade for general use is HB. Virtually all points of measurement, such as the distances from unit to unit of the scales and those of the curve and each bar, should first be indicated on the section sheet by light lead pencil marks or dots. This use of the dots will facilitate the drawing of the lines, curve, and bars in ink. The entire curve and an outline of each bar might with advantage first be drawn in light lead pencil, the ink being later superimposed after the student has satisfied himself that his pencil markings correctly represent the data. The dots and other lead pencil markings can be erased after the ink 70 Chartography in Ten Lessons has dried. It is much easier to correct a mistake made in lead pencil than one made in ink. The curve itself is made finally with the draftsman's ruling pen. The neat lines of the frame are drawn in ink after the framework of the chart has been entirely completed. THE KIND OF INK The best black ink for charting purposes is Higgins' American India. In purchasing ask the dealer for waterproof quality. This, when it dries, is insoluble and will not smear or spread in case the sheet is brought in contact with water, as is often the case when the chart is to be re- produced by the blue-printing process. Another favorable quality of this ink is exhibited in the process of drying areas on charts, such as bars, as it dries with a flat or "dead" surface. Such a surface is highly desirable in case the chart is to be reproduced by such photographic processes as zinc-etching, photo-lithography, and so on. "Chin-chin" ink, also an India ink and noted for its opacity, can be used to special advantage in cases where the chart is to be reproduced by the blue-printing process. While special mention is made of these two inks, there are also other India inks on the market equally as good for ordinary charting purposes. With the smaller bottles is Tools of the Chartographer 71 usually supplied a beveled quill inserted in the cork which is used in filling the ruling pen. THE RULING PEN This ruling pen is an invaluable tool to the chartographer. It has two blades or tines the relation of each to the other being controlled by an adjusting screw. The manipulation of this screw permits the drawing of lines of varying widths. The use of the ruling pen should be con- fined to line and curve work. Some pens have a lever attachment which permits the cleaning of the tines without disturbing the gauge at which they may be set. This lever saves the time re- quired to make the proper adjustment again and prevents the possibility of the chartographer resuming work with a different adjustment of the tines. CORRECT position FOR HOLDING PEN The ruling i>cn should be licld in such a j)osition a.s to be in a plane i)eri)eiidicular to the surface of the drawing sheet, the tips of the thumb and forefinger grasj)ing the pen at the adjusting screw. This is illustrated on the following page. Holding the pen in this way permits, when necessary, the manipulation of the screw by slightly raising the pen from contact with the 72 Chartography in Ten Lessons sheet. In ruling lines or curves the hand or fingers should not touch the paper, nor should the elbow rest on the sheet. The movement of the pen is not from the hand but is a free elbow movement and is from left to right and from bot- tom to top of the sheet. Tools of the Chartographer 73 Failure to observe these instructions will result in the points of the tines wearing away unevenly and the pen then develops what is referred to by draftsmen as a "shoulder." This usually means that this particular pen must be discarded for line and curve work as it is no longer a "true" instrument or tool. These old pens can be used to advantage, however, in filling in bars, they being operated in such cases somewhat as one would a small brush. It is not impossible to remove a "shoulder" from a ruling pen. This can be done by using a small oil-stone or razor-hone. The stone or hone can also be used to advantage in keeping the points of the tines sharp and true. In this process of sharpening be careful to hold the ruling pen against the surface of the stone or hone at an angle of about forty-five degrees, grinding the points with a gentle rotary motion. Follow this by rubbing the points of the tines on any glass surface. .\n examination of the points should then find all "burrs" or unevenness to have been removed. PEN POINTS In addition to the ruling pen and as a substitute for it in many uses, the .student will need pen points and, of course, a penholder or holders. For fine line work and small lettering Gillett's 74 Chartography in Ten Lessons No. 303 is recommended. Esterbrook's No. 14 bank pen point is also good for lettering. Gillett's No. 291 will also be found satisfactory, especially in mapping work. The student is no doubt familiar through per- sonal experience with the fact that most pen points when first dipped in ink repel or throw off the ink. This is likely to result in blots or spots if it occurs on a sheet of drawing paper. To obviate this it is suggested that the pen point be held for a moment in the flame of a match before being put to use for the first time. THE drawing board The effective use of the section paper, the pen- cil, the ruling pen, the pen points, and the ink makes necessary that the student also have a drawing board. This is nearly always made of neatly glued strips of soft wood, usually white pine, with a hardwood ledge of an inch or so on each end. The board can be secured in varying sizes ranging from 12 by 17 inches to 31 by 42 inches. Larger sizes can also be purchased. The board rests unattached on the desk or table and can be moved about freely with the section paper temporarily attached to it by means of thumb tacks. In case the student prefers a drawing table, this can be had in various makes and designs. Tools of the Chartographer 75 the t-square The drawing board or table facilitates greatly the use of the T-Square, another tool of the char- tographer which he will find invaluable. It is so-called because of its resemblance to the cap- ital letter T. For all purposes of accurate line drawing not involving measurement it supplants the ordinary ruler. A T-Square fitted with trans- parent ruling edges is reconmiended, as it permits the draftsman to see adjacent portions of the section sheet that would be hidden if a wooden straightedge were used. It fits in snugly and along either ledge of the board or table accurately by reason of the head piece of the T-Square ex- tending beneath the blade with its ruling edges. This permits of a true base line as well as other horizontal lines. Upon this base line, with theT- Square in position, true vertical lines are erected by means of the Triangle. THE TRIANGLE The use of the Triangle is largely confined to making vertical and horizontal lines. Do not attempt to draw these lines with the ordinary ruler if any degree of accuracy is desired as such an attempt will most likely result in inaccuracy. Accuracy, it should be remembered, is one of the cardinal principles of good chartography. 76 Chartography in Ten Lessons Fia I Fie.a Tools of the Chartographer 77 The illustration on the preceding page shows some of the uses to which the Triangle is put when operated in connection with its running-mate, the T-Square. Triangles are obtainable in numerous sizes and angles, the standard angles being 45 degrees and 30 by 60, the latter commonly called "Thirty" by draftsmen. THE engineer's SCALE Important uses will also be found in chart making for the engineer's rule or scale. It is an equilateral triangle in shape, that is, all its sides are equal; it is usually made of hardwood, 12 inches in length (although different lengths are procurable), and has three edges each with two measuring surfaces. These six surfaces are laid off into multiples of 10, with 10, 20, 30, 40, 50, and 60 units to the inch, and in consequence they provide measurements of almost any fraction of an inch that can l)e quickly applied to varying scale units of less than an inch. The engineer's .scale is admirably adapted to linear measure- ments, that is, to measurements pertaining to or of the nature of a line or in one direction. The engineer's triangular scale is not to be confused with the architect's triangular scale, the latter having tlie inch divided into units of fourths, eighths, sixteenths, thirty-seconds, and so on, 78 Chartograpiiy in Ten Lessons and which is of Httle use to the chartographer. The student is cautioned against making use of the engineer's rule for ordinary ruUng purposes, as this use wears away the ruled edges and in time makes inehgible the sub-divisions of the inch. THE DIVIDERS Assistance in the drawing of a chart is also rendered by the use of the compass or dividers. It consists of a handle from which extends two prongs each having a sharp point. In the handle is a joint, either a pivot or tongue, which permits adjustments between the two points up to several inches. This enables the draftsman to "step- off" or gauge accurately any measurements on lines or charts that are to be transferred to other lines or charts. Greater accuracy will be se- cured from the use of the dividers than from that of the ordinary ruler for this purpose. Every draftsman has use for kneaded rubber, art gum, and the "ruby" or red rubber eraser for erasure purposes. A hard rubber or gritty eraser, such as the ordinary typewriter eraser, should not be used. A supply of pins, clips, thumb tacks, and the like will also come in handy. THE ESSENTIAL TOOLS The following summarizes the more important tools needed in chartography : Tools of the Chartographer 79 One drawing pencil HB. One ruling pen. Six Gillett's No. 303 and six Easterbrook's No. 14 pen points. One bow or compass pen with interchangeable pen and pencil points and extension bar. One bottle Higgin's waterproof black drawing ink. One drawing board or table. One T-Square. One six-inch celluloid 45 degree triangle. One twelve-inch engineer's triangular rule. One dividers. These tools can each be bought separately but a material saving is made by purchasing a com- plete set of drawing instruments at the outset. The price naturally varies according to the quality but an expensive set is not necessary for good work. The outfit of the chartographer may be simple or elaborate according to individual taste. A few well selected instruments of standard make is recommended at first. Where the begiinier confines himself to a limited number of tools he becomes familiar with the "feel" and balance of each instrument and, as a result, soon learns to liandle it with confidence and skill. This applies especially to the drafting or ruling pen. QUESTIONS FOR SELF-EXAMINATION 1. What is cross section or coordinate paper? What ser- vice does it perform in chartography? 2. What is the function of the lead pencil in chart making? 3. What is the ruling pen? Describe the correct posi- tion for holding it. What is a "shoulder" and what causes it? How can it be prevented? 4. Describe the drawing board and its uses. 5. What is the T-Square and what are its uses? The Triangle? 6. What is the engineer's scale? How and for what purposes is it employed? 7. Describe the dividers and its uses. 8. What are the essential tools in chartography? 80 LESSON VII Accuracy in Chartography The Use of the Typeivriter — Drawing Letters for the Title — Exaggerating the Curve — Effects of Exaggerating the Curve — Advantages of Extra Squares. In placing on the chart the figures of the scale lines, the statistical table, the foot-notes, and other figures and letters the use of the typewriter enables the chartographer to meet many of the exactions encountered in the practice of his art. This is especially true when a large number of different charts has to be made for reproduction in quantities by means of one of the photographic processes. THE USE OP THE TYPEWRITER If a long-carriage machine is not available and if the coordinate sheet is too large for the ordinary typewriter the sheet can })o cut in two and after- wards pasted together. This makes necessary the exerci.se of care in handling the sheet after- wards or else the typewritten figures will "rub." This work on the typewriter should be done after the lines and curve or burs are comj)leted. Another j)ractice that has many advantages is 81 82 Chartography in Ten Lessons first to typewrite the numbers and words on separate slips of paper and then paste these securely in their proper places on the section sheet. This plan should be followed if the chart is to be reproduced. It enables corrections to be made more easily and does not wrinkle or otherwise damage the sheet. If the chart is not to be reproduced, the letters and figures should be written according to the first plan, that is directly on the coordinate sheet itself. Typewriting the table directly on the chart or on a separate slip of paper and later pasting this on the sheet, requires considerable painstaking care. The tyj^ed figures and letters must be clean, decimal points separating the digits must be in column order equally exact with the fig ires themselves, units of tens or hundreds and so on must be under each other, and all in straight columns with headings appropriately placed at the top of each. In pasting the slip on the sheet exactness is required so as to avoid the appearance of "skewness." Corrections can more easily be made with the figures and letters on the slip than with them directly typed on the coordinate sheet. In the employment of the typewriter for plac- ing words and numbers on a chart that is to be reproduced by a photographic process, care must Accuracy in Chartography 83 be exercised in seeing to it that the ink of the type- writer ribbon is of a quality that will reproduce. I knoAv of the experience of a fellow chartographer who had in progress a rush contract for several hundred different charts on each of which was to be rei)ro<luced a statistical table. He failed to have pro})er attention given to overseeing the typing of tliese tables, with the result that not one would reproduce because the right kind of ink was not used, In itself the kind of ink may be a small matter but the consequence of not using the right kind is likely to prove serious and costly. DRAWING LETTERS FOR THE TITLE Virtually all the figures and letters required on a chart can be i)laced in their proper positions by means of the typewriter with the exception of the title letters. These lalter are usually larger thaji those of the typewriter, although even for the title the capitals of the tyj^ewriter can some- times be made to serve the requirements. As a general thing, however, the use of the typewriter for the title letters is inadvisable. It is this lettering for the title that is among the exactions of chartograi)hy with which the beginner is likely to have some difliculty. He mu;st learn how to make the kind of letters required. This is not so difficult as might at 84 Chartography in Ten Lessons first appear; in fact it is qujte simple, ahd by a little practice the student can soon become pro- ficient in this phase of the work. In making these larger letters the minute squares of the coordinate sheet are of material assistance. After determining upon the size of the letter required, the horizontal and vertical lines of each letter are drawn by the ruling pen and with the aid of the minute squares. The curved corners are first left blank, as illustrated in the following: mwmi Then the curved corners are filled in with a free hand pen. For the guidance of the beginner in chartog- raphy the letters of the alphabet are reproduced on the opposite page as samples of plain and easily made letters based upon the above instruc- tions as to how they are to be drawn. No attempt is made to present other than a simple utility alphabet, all the letters with the exception U L 86 Chartography in Ten Lessons of I, M, and W being approximately of the same width. Illustration is also given as to the drawing of large figures. EXAGGERATING THE CURVE The beginner in chartograpliy, however, should know how to make letters and should not neglect to become proficient in this direction. Practice in lettering teaches painstaking accuracy, and this is demanded of the good chartographer. In chart making he will have many opportunities for acquiring this personal asset. Especially is this true in the process of determin- ing and plotting the scales for the curve chart. He must be certain that his vertical scale does not permit of the exaggeration of the movepient of the curve. This exaggeration easily results in not allowing for the vertical scale the same amount of space per each scale unit as for the horizontal scale, and vice versa. In other words the movement of the curve can be exaggerated either vertically or horizontally. "The scales of any curve chart should be so selected," says Brinton, in Graphic Methods for Presenting Facts, "that the chart will not be exaggerated in either the horizontal or the ver- tical direction. It is possible to cause a visual exaggeration of data by carelessly or intentionally Accuracy in Chartography 87 selecting a scale which unduly stretches the chart in either the horizontal or the vertical direction." "The beginner in curve plotting and in curve reading," continues Brinton, "is apt to be somewhat puzzled by the different effects which may be obtained bj' changing the ratio between the vertical scale and the horizontal scale. It is difficult to give any general rules which would assist in overcoming the beginner's confusion. Ordinarily the best way to get facility in making the proper choice of vertical and horizontal scales for plotting curves is to take one set of data and plot those data in several different ways, noticing the changes which the different scales selected give in the proportions of the chart. Just as the written or spoken English language may be used to make gross exaggerations, so charts and especially curves may convey exag- gerations unless the person preparing the charts uses as much care as he would ordinarily use to avoid exaggerations if presenting liis material by written or spoken words." "A person reading charts must take great care," concludes Brinton on this point, "that he does not give too much weight to the actual appearance of the curve on the page, instead of basing his conclusions on the |)erccntage increase or decrease as ju<lged from the figures of the ver- 88 Chartography in Ten Lessons tical scale. The proper choice of scales for curve plotting is largely a matter of judgment, and the judgment can be trained very greatly if it is kept in mind to examine every curve chart which comes to one's attention to see whether the verti- cal and horizontal scales have been selected so that the chart gives a fair representation of the facts." EFFECTS OF EXAGGERATING THE CURVE The effects of an exaggeration of the vertical scale can be seen from a study of the chart on the opposite page. The units of the vertical scale are there purposely made twice the distance apart than are units of the horizontal scale. The result is an exaggeration to the eye, in the rise and fall in the movements of the curve across the sheet, of just twice what these movements should be. In order that the student may comprehend clearly for himself just what this exaggeration of the curve means, it is suggested that he draw on the chart on page 89 in light lead pencil a broken or dash curve that conforms to a scale by which the distance between the horizontal lines is re- duced one-half. In other words, he is to give to each horizontal and vertical scale unit exactly the same space. Rearranging the vertical scale units accordingly the unit 7 falls on the vertical line at a point half IMMIGRATION TO UNITED STATES BY MONTHS, 1918 Tboosaiids of Ixalgruiti Jan. Feti. Mair. Ipr. Kay June Jul; Aug, Stpt. 0«t. (OT. Mo. 1« \ Iknth Iiimlgranti Ju. 6,3S« Mb. 7,388 lUr. 6,510 Apr. 9,641 Miy IS. 217 JtUM 14, £47 Jnl^ 7,780 tag. 7,863 sept. 9,997 Oct. 11,771 HOT. 8,499 Dee. 10,748 1 \ 1 L£ i 10 1 t ? \ / / \ / , / ^ / / \ / \ / « 16 U U 11 St«tt>tla« V* lT<m Oaimtu of I^IcrctlOB n, 9. I>«p*rta*Dt of lAbor 90 Chartography in Ten Lessons way between the unit 6 and the present 7, and the point now marked by the latter becomes the place for the scale unit 8. Proceeding upward in this manner cross out with the lead pencil each of the old vertical scale units and substitute the new units according to the revised plotting. Re- produce in lead pencil this new scale also alongside the extreme right vertical line. Connect these new left and right vertical scale points by horizon- tal lines in light lead pencil. Next draw the curve from point to point of the vertical lines as determined by the revised units. It will now be found that this new curve moves up and down exactly one-half the distance of the original curve. The new curve starts at a point on the left vertical line just above the pre- sent scale designation 6 and ends at a point on the right vertical line half way between the present scale units 8 and 9. These vertical scale units represent thousands of immigrants, as explained in the word designation just above the horizontal scale line. It is important to remember in connection with the exaggeration of the curve that the arbitrary limitations of space imposed upon the chartog- rapher does not permit him in every case to choose the scale that might have been chosen if there were no factors to consider other than the exact Accuracy in Chartography 91 presentation of the statistics. His problem is to present the facts as clearly as possible within the arbitrary limitations of space imposed upon him. Sometimes he will find himself in a quandary in his endeavors to include all the necessary data without exaggerating one or the other of the scales. ADVANTAGES OF EXTRA SQUARES Reducing by one-half the space allowed the ver- tical scale unit in the chart on page 89 brings the lowest and the highest points of the curve within a space of less than three inches, thus decreasing horizontally as compared with the old curve the size of the area within which the curve moves without affecting its size vertically. Plotted in this way results in an awkward size for the cliart. This can be overcome by providing at least two series of squares both below the lowest and above the highest i)oints recorded by the movement of the curve. In such cases the vertical and hori- zontal lines forming the squares are drawn just as if they were to be used to indicate a stage in the movement of the curve. This extension of the area of the squares should also be regulated so as to accommodate the placing of the statistical table without crowding. It will nearly always be found feasible in plotting the vertical scale to 92 Chartography in Ten Lessons provide for at least one series of squares vertically into which the movements of the curve do not enter. This will be found to be advantageous in a number of ways. It adds to clearness of expres- sion as well as avoids the appearance of crowding. If the exaggeration of the scale in the chart on page 89 had been in the horizontal instead of the vertical measurement, just the opposite effects to those noted would have resulted. It is suggested that the student draw a chart in which he gives twice the space to the horizontal scale unit that he gives to the vertical scale unit, using the data in the statistical table of the chart on page 89. QUESTIONS FOR SELF-EXAMINATION 1. Describe the various uses of the typewriter in chart- ography. 2. How are large letters drawn by hand? 3. How is the curve exaggerated? What are some of its efifects? How can these be avoided? 4. What are the advantages of extra squares? LESSON VIII Curve and Bar Designations Disadvantages of the Unbroken Curve — Cun>e Designations — Word Designations of Curves — The Peak-Top Curve — Determining the Scale Spacing — Utility of the Curve Chart — Chartog- raphy Based on Comparisons — Bar Designa- tions — Jnterpretiyig the Bar — Some Character- istics of a Good Bar Chart — Word Designation of Scale Units. It is plain that if a number of curves on the same chart are each drawn as an unhroken curve much confusion will accomjjany efforts to interpret the tendencies shown, as most likely the curves cross and re-cross each other. This is illustrated in the chart on the following page. DISADVANTAGES OF THE UNBROKEN CURVE Let the student try to follow each curve from its beginning on the left vertical scale line to its termination on the right vertical scale line. It is hardly possilde that he acconiijlislies the task successfully in every case by ending on the curve he starts out upon, as indicated by the initial abbreviation of the name of the railroad. Even if he does succeed he will have .spent a great deal 98 hi i tc S Ul g b. O oc u o s 3 < c hi > < O f« « f4 A 9t •-« « amy o t^ j j J2 *• ^ £ * • I . B a; S a 3 «*<»:• <i>>« *■<! Curve and Bar Designations 95 more time than should be required to interpret such a chart correctly. Among the aims of chartography is to prevent confusion and to aid comprehension at a glance, and the reading of a chart should not have placed in the way obstacles like those illustrated on the opjjosite page, es- pecially when the obstacles have no reason or even excuse for being. A study of the chart should quickly convince the student of the dis- advantage of using the same kind of unbroken curve for two or more statistical elements on the same chart. Reading the chart in question is only slightly aided by placing at the left and right vertical scale lines the abbreviations of the names of the different railroads which the curves represent. CURVE designations Attention is thus called to a practical condi- tion confronting tlie chartographcr which would be replete with difficulties did he not have re- course to a sinijjle device to overcome them. This is the enij>loyment of difroront designations for two and more curves. In contrast with the utibrokon or straight line curves of the diart on j)agc 04, those of the chart on page 96 should be studied. 'I'he latter com- pare as many as nine sei)arate an<l distinct sta- AVERAGE PRICES'OF MEAT PRODUCTS UNITED STATES, 1913-1919 C«nti 19U XI BACon Zi Sirloin Rirund 3t«aX Pork Chopa l.Amb 20 Onok ItoAet — 191} 1914 191S 1916 1917 )9L« 1919 Bacon >&.' 26.7 26.4 28.1 M.i 49.5 i7.2 Han 26.6 26.6 25.} 31.2 36.5 44.6 52.9 Sirloin Ste«» 26.4 26.4 25.1 27.0 }1.7 36.6 43.7 BounO StMlt 22.3 23.0 22.3 24.0 26.9 34.5 40.5 Porn Chop» 21.4 21.6 19.7 22.5 30.6 36.6 41.4 Umb 20.2 19.3 21.0 23.0 27.6 36.3 39.9 016 fiM«t' 19.9 20.1 19. » 21.0 25.2 29.3 34.6 OIUOK RoMt 16.2 17.0 16.0 21.2 21.2 26.5 29.4 puts B««f 12.2 12.4 12.2 12.6 l«.l 19.9 22.6 / / // 1 / 1 i / / / / / / / / /A 1/ / / / 4 / "/. / ^^^ . y" y'' 840M 45 Sirloin Por* Chop* RouaA St*4ll 35 mil II044I 90 CBuok lout 25 Plato BooC to Stall etico or* froa MooVtly Labor Bavltv, pp. 77. U 5 euroau of Labor SUtlotlca. * Ataraga prico l* of April 13 of aach yaar Curve and Bar Designations 97 tistical elements, presenting two more curves than are in the chart on page 94, and yet there is not the slightest difficulty or confusion in tracing the nine curves from their beginning to their termination. This greater clearness and ease of interpretation is almost entirely due to the fact that a different designation is given to each curve. If this had not been done it would be almost as difficult to follow the curves in the chart on page 96 as in the chart on page 94 — the curves of the former would also be lost, as to the reading of their movement, at the points where they cross and re-cross one another. Quite frequently the student of chartography will encounter the problem of having many curves to compare. While this difficulty is met in part by employing different designations for the curves, there will be occasions when even this method will result in confusion. Under such circumstances, instead of attempting to draw all the curves on a single chart, it will be found advantageous to make two or more charts. One set of the group of figures should be .selected as a common basis for the comi)arison and the curve re[)resenting this set or statistical element in- .serted on all the charts, this curve taking the .same unbroken or straight line designation on each chart. It i.s a mistake to j)lace a larger 98 Chartography in Ten Lessons number of curves on one chart than can be read quickly and without confusion to the eye in tracing their movements. Where the curves lie close together or are constantly crossing and re- crossing each other, more than five or six are likely to result in this confusion. It is in making clear just such problems as those presented in the chart on page 96, where a number of different statistical elements must be compared, that the advantage of the curve method over the statistical method becomes apparent. To grasp quickly and comprehendingly the mean- ing of each of the nine different columns of figures, not only in relation to its own element over the period of years but also in relation to each of the elements of the other eight columns, is prac- tically an impossibility to most minds. And yet one of average intelligence can easily read the trend or tendency of these prices of different kinds of meats when interpreted by the curves. WORD designations OF CURVES Not only from the point of view of interpretation but also of mechanical construction the chart on page 96 is recommended for close study. Note the word designations of the curves to the left and right of the vertical scale lines. This inser- tion of the word designation alongside the point Curve and Bar Designations 99 of contact of the curve with the vertical scale lines lends to easy reading of the chart. It requires, however, extending the space between each vertical scale line and its respective neat line, and this is not always possible. In such cases the curve designations with their word descriptions should be placed at some convenient place on the framework itself as a sort of key or legend. Quite frequently a good place for the legend will be found to be just beneath the bottom horizontal line and above or between the foot- notes. The different designations that can be em- ployed for curves should be practiced by the student until he has acquired facihty in drawing them. To assist him in this the following page of designations is presented. These have been made considerably larger than is necessary for the curve on the chart. THE peak-top curve The student is cautioned against making use of what is called the "stairway" curve. This makes a flat or step-like c-ontact at tlie point <letcrniined by the scale. All curves, as has btieu said, should approach the j)oint of contact slantingly and direct from the j)oint previously touched on the vertical line. The great advantage of this Curve and Bar Designations 101 peak-top curve is brought out on charts contain- ing two or more curves which approach each other at or near the same points. In such cases the peak- top permits of easy separation of the two curves and does not result in confusion caused by in- ability to follow the curves, which latter is inevit- able when two flat-top or stairway curves ap- proach each other at the scale unit points. DETERMINING THE SCALE SPACING In determining the vertical space a number of curves should occupy every one of the columns of figures in the statistical table is examined to ascer- tain the lowest and highest numbers that are to be charted. That is, for this purpose all the differ- ent columns representing the comparable statis- tical elements are considered as if they were only a single column. Take, for illustration, the table of the chart on page 96. The lowest number of cents represented in any one of all seven columns is 12.2 in the col- umn for the year 1913. This number also api)cars in the colunui for the year 1915. The largest nuinlx^r of cents in any one of all the columns is 57.2 in the column for the jear 1919. Thus the spread for all the numbers in all the columns is from 12.2, the j)rice of plate beef in 1913 (also in 1915), to 57.2, the price of bacon in 1919, or a max- 102 Chartography in Ten Lessons imum spread for all the figures of 45. With a verti- cal scale unit of 5 this spread requires at least ten squares vertically with the base line starting at the unit 10. Starting at would require two addi- tional squares below the scale unit 10, but if this were done there would not be space enough any- where on the framework for the inclusion of the statistical table. Our squares are thus more valuable at the top, so we provide two additional there to accommodate the table, UTILITY OF THE CURVE CHART By this time the student should have become impressed with the great utility of the curve method in chartography. In comparing the tendency over a period of time of two or more dis- tinct but, related statistical elements it is far superior to the bar method in chart making and incidentally is also superior to the statistical me- thod. While a trained statistician could interpret satisfactorily the tendency from a study of col- umns of figures, no one else could perceive the movement as clearly as it is convincingly demon- strated by curves drawn in relation to each other. This is particularly true when more than two curves representing different columns of figures are compared, as in the chart on page 96. These curves not only show the variations in each of the Curve and Bak Designations 103 items for each year compared with the other years and with the other items but they also give a comprehensive perspective of the entire movement during all the years; they show the status of each of the items in relation to every other item for each year and at the beginning and through to the end of the period of time. Thus it is that chartography can be said to "speak" a language easily made intelligible to the mind through the eye. CHARTOGRAPHY BASED ON COMPARISONS The curve also strikingly emphasizes the fact that the art of chartography is based upon rela- tions or comparisons. There can be no chart without a comjjarison of some kind. And as the very nature of statistics involves a relation be- tween or a comparison of groups of facts, it is not too much to say that chartography is the art best adapted to expressing this clearly and con- cisely. "Comparison is, in general, the final goal to- ward which all statistical studies tend," says King, in his Elemenls of Siatifitical Mclliod. "Com- parison is necessary to give us clear ideas of the relationship of things in time and space. It is also essential in determining whether phenomena are connected or independent and in establishing relations of cause and efifect. We may wish to 104 Chartography in Ten Lessons study: 1. Changes of a single variable. 2. The structure of different groups. 3. Changes in two or more variables." Brinton, in his Graphic Methods for Presenting Facts, puts it this way: "One of a business man's chief assets is his ability to show things to others in their true proportions. He is continually mak- ing contrasts, and holding up for comparison different propositions which come up in his daily affairs. The graphic method lends itself admir- ably to use in making comparisons. It is surpris- ing how much clearer even simple comparisons of only two or three items will appear when their numerical value is put in graphic form rather than in figures." In every chart, then, a comparison or relation of some kind is involved. This comparison may be with the same statistical element for two or more periods of time; it may be of two or more differ- ent elements for the same or different periods of time. It may be a comparison of total or abso- lute amounts, of increases or decreases, of the rate of change. It may be a relation of one or more elements expressed in proportions to a common total, and so on. Graphic methods must, of course, show compar- able facts only and these in their true relations and profKjrtions in order to present the correct sit- Curve and Bar Designations 105 uation. They represent the best kno'VNii scheme for presenting contrasts and in this way indelibly impressing their significance upon the mind. Hav- ing recourse to curves differentl}' constructed permits some of these comparisons to be made much more clearly than if the chartographer were limited to the unbroken or straight line curve. bar designations DiflFerent designations for different statistical elements or factors apply with equal significance to the bar as to the curve chart. The simplest designations are plain black and white which are usually employed where only two groups of figures or statistical elements are involved. The plain white bar is secured simi)ly by outlining it on the section sheet and without filling it in with the ink. But it is not as satisfactory as a cross- hatched bar, that is, one with the outline filled in by drawing liglit diagonal lines. One use of designations in bar charts is illus- trated in the chart on page 106. The student should write out on jiajjcr a careful analysis of this chart, not only from the point of view of its construction but al.so from that of interi)retation. Another and probably the most common use of designations in bar charts is illustrated on page 107. It shows the employment of the black OPERATING REVENUES AND EXPENSES PENNSYLVANIA RAILROAD UlUoM of BolUn 100 us 1*0* ino i«u itu itu IfU ins iti« •M «Vl » n «« trm Mpert* of R&llroat to lat«r»taM eoBwo* CooniMioB < 10 K s I t i 1 g I < • 1 8 •• V) I i I II O •I t3 108 Chartography in Ten Lessons and white as parts of a whole bar. In this chart the total freight traffic of the Norfolk and Western Railroad Company has been separated on the percentage basis between bituminous coal and all other commodities transported. The com- parison involved is expressed as a ratio. The changes over the period of years of the coal traffic in relation to the total traffic, as well as also in relation to the traffic in all other com- modities, is seen by comparing the black portions of the bars with the total bars, reading from left to right. Similar changes in the proportion of the traffic of other commodities to the total freight traffic is shown by comparing the white portions of the bars with the total bars, reading from right to left in order better to secure an idea of the differences in the length of the white sections. The proportion of each designation to the total bar in any one year and the changes as between years are clearly indicated. In this chart each of the series of horizontal bars represents by 100 per cent the total amount of all freight traffic for each of the designated years. Consequently all the bars are of the same length. No information is given as to the numbers representing the absolute amount of traffic of the road, which it may naturally be assumed varied in the different years — it may Curve and Bar Designations 109 have increased or decreased from year to year and if these numbers were charted they would likely give bars of varying lengths. All that the chartographer is interested in, so far as the sta- tistics enlighten him, are the changes in the pro- portion of the two components which together comprise the total freight traffic for each year. SOME CHARACTERISTICS OF A GOOD BAR CHART The chart on page 107 is a good illustration of this kind of a bar chart. The title is concise and yet comprehensive. The sub-title — that of the particular railroad — is well placed and well spaced. The figures for the years are directly under each other and are spaced sufficiently from the bars, while the figures for the black and white portions of the bars are directly in proper column form on each of the six bars. The reader is told at the top of the framework tliat the figures rej)resent per cents, and at the bottom among the foot-notes that the statistics have been "Com- j)iled from Hei)orts of the Railroad to Interstate Commerce Conunission." The group of bars is .so spaced as to avoid the appearance of crowding. The designations of the bars are clearly distin- guished between " Bituminous ('oal" and "Other Commodities" by' means of the legend beneath the bars. The neat lines |)roperly frame tlie series of bars. a: < o z g a: lu u. O o < 0. < o o < z z o IJL o z o I- m E ♦- o e8 hJ *^ m Jl 2 K wi iii ii • ----- 3 *H _ . - . . o 1 II II Curve and Bar Designations 111 When more than two statistical elements have to be indicated on a bar the chartographer has recourse to an almost unlimited number of differ- ent designations. Some of these, shown in the chart on the opposite page, indicate the extent to which variation in bar designations can be carried. The effect of the use of these different designations is for the purpose, of course, of causing the areas to stand out in contrast with each other. The student should practice until he becomes proficient in making these designa- tions. In the chart on page 89 (Lesson VII) and in the one on page 106 will be found a simi)le device and yet a valuable aid to chartography. This is the word designation of the vertical scale unit of a curve and the horizontal scale unit of a bar chart. In the chart on page 89 (Lesson VII) this designation is "Thousands of Immigrants." By its use the chartographer is al)le to droj) from each of the vertical scale units all the ciphers repre- senting thousands. That is, instead of the vertical scale starting with the unit 0,000 it })ogiiis with 6, the interpreter of the chart knowing from having read the word designation that this unit G means 6,000. So with the horizontal scale unit of the chart on page 106. There the word designation is "Millions of Dollars" and this permits the drop- 112 Chartography in Ten Lessons ping of six ciphers from each of the horizontal scale units — it allows the scale to begin with the unit 100 instead of requiring 100,000,000. To reproduce the additional six ciphers after each one of the horizontal scale units would over- burden the horizontal scale line with figures even if s]iace could be found for all of them. Making use of the word designation of the scale so as to drop the ciphers from the scale line itself has many advantages and should always be employed where the numbers represented are more than three digits, that is, thousands and over. The designation should be clearly pre- sented in an easily observable place on the chart, even when the table of figures would seem to make this unnecessary, so that the chart can be quickly read and easily understood without reference to the statistics from which it is made. Usually the best position for the word designation is just beneath the top neat line and centered above the horizontal scale line. QUESTIONS FOR SELF-EXAMINATION 1. What are the advantages of various designations for different curves? 2. Describe the peak-top curve. 3. How is the scale spacing determined? 4. Describe the utility of a curve chart. 5. What is the basis of chartography? 6. Discuss the uses of various designations for different bars. LESSON IX Value of Statistics to Chartography The Statistical Table — Aids in Reading the Table — The Stibsfitntion of Ciphers — The Table of Ratios— Building Up A Table— The Percentage Increase and Decrease — The Zero Line — The Arithmetic Average — The Misuse of the Average — Statistical Class Limits. It should be clear by this time that a most important asset in the practice of chartography is a knowledge of statistics. The mere mechani- cal act of drawing lines and curves and bars on section paper is the work of the draftsman and does not of itself make a chartographer. In these Lessons it has been assumed that the infor- mation is already at hand in proper form for the drawing of the chart — that the collection and comi)ilation of the statistics have been correctly done and that the figures have been checked and verified so that there is no question as to their accuracy and trustworthiness. This assunii)tion has l)een necessary for the reason that statistics are a distinct field of study in themselves, with sulj-divisioris as to methods of collection, of comj^ilatiori, of tiibulation, of comjju- tation, of arrangement, of presentation, of inter- ns 114 Chartography in Ten Lessons pretation, and so on. This field is entirely too extensive to be presented in these Lessons even in merest outline, and for a knowledge of its principles the student should have recourse to standard books on the subject. All that can be done here is to make the briefest reference to a few only of its features which most vitally concern the beginner in chartography. THE STATISTICAL TABLE Conspicuous among these is the arrangement of the statistical elements in the table. The prin- ciples underlying this have necessarily been dis- cussed briefly in preceding Lessons. But there is one other point in particular to which attention must be called. This is the more or less common practice — fortunately it is becoming less common as the rules of good chartography become more widely disseminated and better known — for charts to appear in otherwise first-class publica- tions with the statistical table having the latest date at the top and with the earliest period of time at the bottom of the column. A chart made from a table arranged in this way reads backwards from the latest to the earliest year. In order to interpret it from the earliest year and in sequence of time it has to be read from right to left, which is the wrong way to Value of Statistics to Chartography 115 read a chart. Invariably this arrangement is at first glance misread, as the natural inclination of the reader is to assume that years are arranged in proper sequence. Where they are not so ar- ranged too much time is lost before this is realized. The first impression on the interpreter of the chart in such cases is exactly the reverse of that intended and which would have been received by him if the correct method of arranging the statistical elements in table form had been fol- lowed. In consequence, one of the fundamental purposes of the chart method of disseminating knowledge is violated. Such a practice should not be indulged in even in exceptional cases. RECONSTRUCTING THE TABLE Whenever the chartographer has a table of figures to chart in which the latest year or period of time api)ears at the top of the column, he should rearrange or reconstruct the table in correct col- umn form with the earliest year at the top before he begins plaruiiiig his chart. The justification for presenting the latest year first in the column is that it is of greater impor- tance compared with the other years recorded. As a matter of fact, no one year in a series of years that charts a tendency is of any greater impor- tance than the other years. It may be of greater 116 Chartography in Ten Lessons importance in the mind of some particular indi- vidual or individuals as to the significance of the data it discloses but in itself as a year it is only of equal importance with every other year. Besides, chartography offers other and much better methods for placing emphasis on particular statis- tical elements. AIDS IN READING THE TABLE It may be advisable, where large numbers are the basis of a chart, to substitute ciphers as the last two figures in tens of thousands, the last three in hundreds of thousands, and the last five in millions. By raising or lowering the last preceding digit before the cipher, sufficient accuracy is obtained for the purpose of most cha ts. The advantage of this is that the ciphers enable the mind to grasp more quickly the signifi- cance of the numbers. In financial statements, however, this practice has objections. Necessary space on the chart for the columns of figures can sometimes be secured by dropping entirely the last three or six digits and substituting at the head of the columns a word description of the amount represented, such as thousands or m illions and so on as the case may be. This elimi- nates the confusion to the eye of numerous digits. While this practice is not only admissible but also advisable in designations of the scales it should Value of Statistics to Chartography 117 not be allowed to become general in tabulations, not even where the value or volume or quantitj'^ is so large as to run into seven or more digits, as long as there is space on the chart to show the complete numbers without crowding. Assistance in the direction of facilitating the rapid reading of a statistical table and in enabling a quicker grasp of its significance, is rendered in cases of long columns of figures by breaking up the numbers into groups of fives with double the space separating each group. For illustration, instead of presenting the table on the chart like this: Year Population 1900 75.994,575 1901 77,747,402 1902 79,365,396 1903 80,983,.390 1904 82,601,384 1905 84,219,378 1906 85,837,372 1907 87,455.366 1908 89,073,360 1909 90.691,354 1910 92,309.348 1911 93,927,342 1912 95,545,.S36 191.S 97,163,.330 1914 98,781.324 1915 100.399.318 1910 102,017,312 118 ClIARTOGRAPHY IN TeN LeSSONS It might better be presented as follows: Year Population 1900 76,000,000 1901 77,700,000 1902 79,400,000 1903 81,000,000 1904 82,600,000 1905 84,200,000 1906 85,800,000 1907 87,500,000 1908 89,100,000 1909 90,700,000 1910 92,300,000 1911 93,900,000 1912 95,500,000 1913 97,200,000 1914 98,800,000 1915 100,400,000 1916 102,000,000 These figures represent the population of continental United States as reported by the Bureau of the Census of the United States Govern- ment in its bulletin on mortality statistics for 1916. Using them as a basis, the student is instructed to draw with a lead pencil a curve or bar chart, whichever he determines, applying to his task the instructions he has received up to this point. Value of Statistics to Chartography llf) THE SUBSTITUTION OF CIPHERS This substitution of ciphers for other digits in large numbers does not affect the accuracy of the chart for the reason that the change made by it in any number is so sHght, compared with its total, as to be lost in the results of the appUca- tion of the scale unit to its measurement. Be- sides, it has a distinct advantage in that it does not accord to the statistics any greater import- ance than the method of their compilation war- rants. No one who is familiar with the methods of taking or enumerating the decennial census of the population of the United States and of esti- mating its growth for intermediate years believes for a single moment that this population, say, in 1916, was exactly 102,017,312 to the last digit of accuracy. While no criticsm of these methods is here intended, it is bul stating the fact that they are not so perfect as to record to the last figure the exact population. Most statistical tables at best are api)roximations and do not represent absolutely accurate and indisputable facts to the point of minute measurement. All that can be expected of chart()graj)hy is that it indicate clearly the general trend or tendency of statistical elements. The chartographcr should be on his guard against i)ermitting the curve or bar to convey an impression of a greater degree of 120 Chartography in Ten Lessons accuracy than is warranted by the statistical information. THE TABLE OF RATIOS In deaHng with tables of ratios it is always advantageous to carry the digits at least one place beyond the decimal point. The advisability of this should be so clear as not to need to be dis- cussed. If, for instance, by not including the last digit of the two ratios 67.6 and 32.4 of the 1912 bar in the chart on page 107 (Lesson VIII) each portion of the whole bar falls short of its correct measurement and the total 100 per cent bar is incomplete. It is hardly ever necessary to carry the digits further than two decimals and in most cases one digit beyond the decimal point answers all practical purposes. Frequently decimals exceeding one-half may be raised to a whole number and under one-half lowered to a cipher. Where the digit is exactly one-half whether it is raised or lowered will depend upon the particular circumstances. Per- centages or ratios are a problem in mathematics and it is to that science the student should have recourse for more complete knowledge of the principles involved. As percentages or ratios are derivative figures it is of advantage, if it does not over burden the chart, to place also on the fj-amework the Value of Statistics to Chartography 121 original figures from which they are derived. Where a choice as to exclusion has to be made it should be in favor of the retention on the chart of the derivative figures upon which it is based. BUILDING UP A TABLE Both the original and their derivative figures appear in the table of the chart on page 122. This chart also illustrates the use of the ciu-ve in expressing ratio. The problem is to show the income of the Pennsylvania Railroad Company in relation to its securities, that is, the rate earned by the latter for each year from 1909 to 1916, both inclusive. The basal information as to securities and income was secured from the annual reports of the company filed with the Interstate Commerce Commission in Washington, D. C. The amount of bonds representing funded debt and the capital stock were added to ascertain the total of all securities for each of the years. This gives the second column of figures in the table to the chart. In order to ascertain the total income that is f)roperly related to these securities the amount of interest paid on funded debt was added to net corjiorate income for each year, which gives the third column. This in- formation enables us to ascertain the rate of income each year by dividing the amounts repre- RATE OF INCOME ON RAILWAY SECURmES i«i» PENNSYLVANIA R. R. ini ml iMs itit / / / / 1 1 \ \ f 1 1 \ / i 1 \ /' IJiMMMk s^ / / 1 1 ! s ^^ \ / / / / / / \ N, N 1 / \ ^^ orti2«2^. . X f / ' / \ — X — — M — — ttrttM lnn> ud Far 0«pl«»l K t Corp«r«i* ttr '««- IMlvtKU iBWrwtA OnA Stock taooa* o«in — l*» 6M,«a3,MD »3,4«.6T« t.M H4.4M.BO0 9.«M,3ll 4.H 1710 T06,979,TTJ 62,2^,600 T.41 411.296. 974 O,04»,l« t.Tl 1911 ri«,SCa,m 4ft, 104.606 «.I4 Ut,443.93B 4,6U.023 7. 47 1912 T14,6ai,60T 4T,4»,U& 4.44 4M,«77.900 T.NU.eSl ft.U inj ra9.0l9.ti4 S0,4U.S11 4.«I 4»i.soa,4n U),6M.4TS e.u ItU T4S.10SJ1B 48.947.004 4.11 4»9,t66.70O l6,3fl7,TT8 7.H ^ in* •U.Ul.aTO 43.41t.su t.U 49».203,600 - n,u«.ui 4.64 h™ in« 9n,us.sT4 a.4«4.'9< t.ti 4M.n4.Toa 1 I3,733,43t I0.»t • e> r<adM MM 1 1 1 1 ! Value of Statistics to Chartography 123 senting capital obligations into the amounts representing net corporate income and interest for each of the corresponding years. The result is the fourth or ratio column of the table. The other column of ratios — the seventh of the table — is ascertained by dividing the amounts repre- senting capital stock into the amounts represent- ing net corporate income. It should be noted that the only comparison made by the two curves of the chart is based upon the two columns of ratio figures. It would better assist comparison in the statistical table if these two columns were placed alongside each other, but this cannot be accomplished with clearness in reading without making another table with the column of years duplicated and with awkward headings above each ratio column. This latter is due to the words expressing the exact meaning of the ratio column taking up a great deal more space in width than is occupied by the three digits and the decimal point. Thus to place adjacent to each other the columns that are compared is likely to commit an ofFense more serious than is the vainc of tlic advantage to be gained. the pp:rcentage increase or decrease Quite a different percentage chart is presented on pa^e 124. It shows the rate or per cent of in- PRINCIPAL ITEMS OF OPERATING EXPENSES. 1008-IOiS BAL-nMORE » OHIO Pl^orM r*pr«a«at ^ar ctnti. laor^a** or Atortvm It ov«r I90ft i«ot tuo \ni lau 11U B<Jl> 41>d TIM rui Pr*l«kt Cu-a Pftaaanflvr Cara ■ 8.M • 9.5T • ».92 "10.93 •IT. 19 •Ji.ia •a.M 42.(9 <T.*6 «i.ta 43. ai U.ll 'U.BO It. 43 ^4.M 11.30 19. Tl U.«0 14.34 • 1.34 • B.«3 ■ 3.96 • t.9» sa.KS so. 23 12.96 22. TS 10. Tt 22.53 ' 3.06 6.40 -f U>oow>ttv«» * Mill* «nd n«a Puel Freight Car* * poadvay tni Trftck Pa«a«in0*r Cart* CmtMM Cnm K«^rt« of liilr<M4 'Ropairt, r«fi««tti9. an4 do9recl«tloo Value of Statistics to Chartography 125 crease or decrease. This arithmetical principle is a most valuable aid to the chartographer, for without it many curve charts that are now possible could not be made. This is true in most cases where the difference between the numbers to be com- pared is considerable, that is, where some are large and the others small numbers, as a chart of these absolute amounts is impossible owing to the requirements of space necessarj^ to indicate the "spread" between the highest and the lowest. But by resolving these \'urying amounts into per cent increases or decreases, as the figures deter- mine arithmetically, statistical elements are secured which permit of a common measurement and in consequence of a comparison of relative movements over a period of years. The percentage increase or decrease in our chart of each element for each year from 1908 to 1913 is based upon the amount for 1908. In consequence every curve starts from the same point at zero on the 1908 vertical line, because it is plain that there could be no increase in 1908 over 1908. As the per cent figures for every element show a decrease in 1909 over 1908 every curve extends downwarrl bellow the zero line to the respective points of contact with the vertical line for 1909. For each of the following years the curves separate more or less widely according to the 126 Chartography in Ten Lessons tendency as indicated by the figures for the different elements. Taking the year 1913 for illustration, there is an increase over 1908 in every one of the prin- cipal operating expense items. It also shows that the expense of buildings, fixtures, and so on had increased faster relatively than that of steam locomotives, or of rails and ties, or of wages, and so on; that the expense of passenger cars had increased less rapidly than that of any of the other items, and that this expense was greater in 1913 than in 1908. THE ZERO LINE The chart on page 124 presents also the zero line feature of chartography based upon percentage increase or decrease figures and not on absolute amounts. It is one that the student is likely to have frequent occasions to make use of. This line is indicated by the cipher designation on the vertical scale lines. The zero line is in reality the base from which the curves move up or down as the numbers of the statistical table and the scale units determine. It practically represents the amounts of each of the eight elements in the year 1908 as indicated by the line drawn horizontally across the chart, for the increase or decrease is "over" that year or line. In other words, the movement of the Value of Statistics to Chartography 127 curves for any one and all of the six years in rela- tion to the zero line is determined by the rela- tion of the amount in each year to the amount in 1908. Thus the fluctuations in the curves from year to year should be read or measured from this zero line and not from the slopes of the curves themselves. Facility in tlie interpretation of such a chart is aided if it is clearly indicated that all the move- ments of the curves above the zero line mean increases over the base year and below that line decreases compared with that year. This is accom- plished by inserting the Mords "Increase" and "Decrease" alongside the vertical scale lines on either side of the zero line. This shows clearly that the vertical scale reads upward from the zero line for increases and downward for decreases. Assistance in the clear interpretation of such a chart is also rendered by making the zero line slightly heavier or wider than the other horizontal lines connecting at other units of the vertical scale lines and at the same time not as heavy or as wide as the curve or curves themselves. This wider zero line calls the reader's attention to the fact that he must interpret the movements of the curves from the zero and not from the lowest or ba.se line. In cases where the figures to be charted show 128 Chartography in Ten Lessons no decreases and in consequence it is not necessary to extend the curves below the zero line, then this line becomes also the lowest or base line at the bottom of the chart and all the movements of the curves are above that line. In such cases it is not necessary to employ the terms "Increase" and "Decrease" above and below the zero line. Nor is it necessary in such cases that the zero line be made wider than the other horizontal lines. While the chart on page 124 designates with a cipher the horizontal line from which the move- ments of the curves are measured, as a matter of fact this line is not a zero line but a 100 per cent line. This is true arithmetically for the reason that in reality it represents the total amount of each element or item for 1908. These were taken as the base from which the figures for each of the other years were ascertained. Arithmetic exactness requires that this line be designated as a 100 and not a line. But in this case chartog- raphy takes liberties with arithmetic for the sake of securing greater clearness in interpretation. Experience has taught that because of the general lack of knowledge on the part of many of those for whom charts are prepared, confusion leading to misinterpretation, and this to misinformation, results whenever the 100 per cent designation is employed in place of the cipher. Value of Statistics to Chartography 129 THE arithmetic AVERAGE In the chart on page 124 the basis upon which the respective percentages have been computed is, as has been said, an amount for a single year. Wherever possible this basis should be the average of the amounts for a number of years, and this is nearl}^ always feasible when the number of years in the table comprises ten or more. This average is ascertained by adding the amounts for the years selected and dividing the total thus obtained by the number of these years. The percentage in- crease or decrease is then computed for each of the years from this average amount. This is not advisable for the statistics in the chart on page 124, as tlie numbei- of years is only six. In those cases where it can be done there will likely be found a material difference between the move- ments of curves over a period of years thus dis- clo.sed compared with the movements shown with only a single year as the base. The advantage of taking the average for a number of years as the base for computing in- creases or decreases is found in tlie fact that this average smooths out the irregularities of high and low or of large and small amounts which may have been comjiriscd in the difrcrent years. If a single year only is used as the ba.se it may be that in that particular year unusual influences or 130 Chartography in Ten Lessons forces were at work to change unduly its total in comparison with preceding or following years and in consequence it is out of normal relation to the amounts of the other years. An illustration of this as to many phases of railway operation, for instance, is the fiscal year 1908 extending from July 1, 1907, to June 30, 1908, both inclusive. The records for that fiscal year include the effects of the panic in the latter part of the calendar year 1907. The railroads were very seriously affected by this disturbance in business and financial conditions and their traffic and revenue fell off strikingly. In con- sequence, any comparison of the operations and finances of subsequent years based upon the single year 1908 would show tendencies that might not and would not be showTi if a year that did not record a panic was used as the base. Averaging a number of years escapes this possibility of statistical error and in consequence avoids misrepresentation in chartography. The proper use of the average is an important asset to the chartographer. This average reper- sents or indicates the usual or common occurrence or status. It is, primarily, as has been stated, a problem of arithmetic. Quite often, if not always, it is simply an arithmetical standard, non-existent in actual reality and yet one around which other Value of Statistics to Chartography 131 facts tend to approximate or conform and by which they are measured or compared. Such, for instance, as the average height of men, or the average price of a pound of bacon, and so on. THE MISUSE of THE AVERAGE The average can be as much of a sinner when improperly made use of as it is a saint when projjerly employed. To the chartographer the use of the average has its pitfalls against which he must be constantly on his guard. While it is indispensable at times, it has its limitations and shortcomings and these must be known if he is to make the best use of it and not be inveigled by its attractions into grievous errors. Quite freqiieiitiy the average comi)riscs ele- ments radically different from each other whose irregularities or dissimilarities have disappeared or been smoothed out to such an extent that it does not represent any measurable status or even approximate situation of the actual facts, and in consequence can have no other effect than to mis- lead. This is illustrated, for instance, in statistics giving tlio average amount of stock hold j)er stockholder in the railways of the I'nitod States. In 1914 this average was stated as .^L'^O.^H. It was obtained by dividing the total number of stockholders — ()':^2,284 — into the total i)ar value 132 Chartography in Ten Lessons capital stock outstanding— $8,685,764,000. Of course, such an average is absolutely meaningless. It is merely an arithmetic average. No such amoimt of stock approaches even in the slightest degree to the actual facts in the case. The fallacy in any practical use of this average can be demonstrated by a simple illustration from almost any railroad. Let us take the Wabash for an example. In 1915 a single stockholder — the Equitable Trust Company — owned $28,744,000 of the stock of this railroad. With nine others, these ten largest stockholders together held $59,449,200 of the stock, or more than sixty-four per cent — nearly two-thirds. In view of these very large single holdings of stock by a very small number of stockliolders — these ten owning by themselves an average of $5,944,920 — any arithmetical com- putation representing the average amount held by each stockholder cannot fairly represent the situation as to the ownership of stock in the Wabash Railroad Company, THE STATISTICAL CLASS LIMITS In such cases as this instead of making use of a meaningless average there is the possibility of recourse to a separation of a group of figures into class limits in order that the facts of a given situa- Value of Statistics to Chartography 133 tion may be more accurately presented. This should always be taJcen advantage of whenever possible. Applied to the preceding illustration it simply means the separation of the total number of stockholders into groups or classes according to selected amounts of stock held. The first step in this statistical process is to determine upon the limitations for the different classes. These are purely arbitrary. To obtain them, round numbers are most desirable, as these give clear cut groups. These numerals form what are technically known as boundary lines of the classes. The difference between them is called statis- tically the class interval. These class intervals should all be equal or uniform. The number of clas.ses and the numl)er in each class become statistically what is called a frequency tal)le. By thus dissecting the statistics a number of verj' interesting and highly important facts is usually disclosed which the j^resentation of the average does not indicate as being j)resent. A knowleflge of these facts is es.sential to a correct and comi>Iete presentation of the actual situation. These facts indicate that the chartograjjlicr must excrci.se his best judgment in the i)resen- tation of the average. He cannot be permitted 134 Chartography in Ten Lessons to excuse himself by hiding behind statistics. Charts that reflect inaccuracies and irregulari- ties of mathematical computation to the extent of being misleading cannot be explained away because, in truth, they never should have been made. This is a high standard to attain, for quite often the chartographer must depend almost entirely upon his statistics for a truthful presen- tation of the facts, and if the statistics are faulty it seems rather unfair to hold the chartographer to strict responsibility for any misleading result. Nevertheless, the chartographer should be as scrupulous and as exacting in the use of statistics as in the use of the English language in maintain- ing a high standard for truthfulness and exactness. A sufficient variety of curve and bar charts have been presented in the preceding Lessons to impress upon the student of chartography that his most important task is the planning of the chart. It should be done before he touches pen to paper in beginning the drawing of the chart. With this planning successfully accomplished the remaining details of the work of execution or construction becomes a relatively simple matter. In this planning the student should first know thoroughly the real meaning or significance of the table of statistics he is to chart — he must "see" clearly the vital point of comparison the Value of Statistics to Chartography 135 chart is to bring out. It is of advantage in com- prehending this point if the student will roughly sketch several different charts, both curve and bar, before he attempts to lay out the curve or bar in ink. He will be surprised at the difference in results shown by the various methods, and can then select the one which best illustrates the significance of the statistics. More time rela- tively should be given to the planning than to the actual drawing of the chart. The time devoted to the latter will be greatly shortened if the planning has been done correctly. Be- sides, it will also save time lost tlirough changes and alterations usually made necessary where the planning has been neglected. QUESTIONS FOR SELF-EXAMINATION 1. What is the value of statistics to chartography? 2. What is the statistical table? How should it be arranged? 3. Describe some of the technical aids to the interpreta- tion of a table of statistics. 4. What are ratios? How are they computed? How arranged in table form? 5. Describe the construction of a statistical table. 6. What are the important differences between ratios and percentage increases or decreases? 7. Of what value to the chartographer are percentage increases or decreases? 8. Describe the zero line and its use in charts showing percentage increases or decreases. 9. What is the average? How is it computed? Describe its advantages and disadvantages. 10. What are statistical class limits? What are boun- dary lines of the classes? What is the class interval? A frequency table? 1S6 LESSON X Primary Principles of Chartography Planning the Chart — Importance of the Right Method — Essentials of Good Chart Making — Planning the Size of the Chart — Planning a Reduction in Size — The Reducing Glass — The French Curves — Checking up the Chart. Chart-making for commercial and other pur- poses is still in its infancy and in consequence has not yet been systematized. Statisticians who are employing it to an increasing extent as an aid in the presentation of facts are in disagree- ment, or rather are not in accord, as to the superiority of different methods. Give a group of statisticians who are also familiar with chartog- raphy a set of figures to chart and^ there will likely be as many fliffcnMit kinds of charts widely divergent in methods as there are statistic;ians. Thus the same information will be charted in many different ways. While variety in charting is possible where numerous illustrations must be made, at the same time some methods are better than others in bringing out the facts more clearly. Variety of effect is permissible and sometimes desirable in order to avoid monotony in presen- tation and to retain attention. 187 138 Chartography in Ten Lessons a choice of methods The value of the chart method being in ex- pressing clearly the meaning of a statistical table, the problem of the chartographer is to select the one method from among the many that will best express this meaning. Particularly is this true in the use of charts by the large corporation. Pre- pared usually for the executive whose time is limited and of great value, the chart must be so drawn as to give to him instantly the true sig- nificance of a mass of statistics which he has not the time to analyze in their details. The efiicient and successful executive must decide quickly and of course correctly. The information furnished him on the chart must not only possess an ac- curate background but it must also be presented to the best advantage if he is to make the right decision and avoid "guess work." The busy executive is more and more being compelled to place greater dependence upon the chart analysis of statistics. As has been made clear, in the practice of chartography there is a choice of widely varying methods. No general rule can be given for determining which of these methods is the best for any particular purpose, but practice will enable one to form his own judgment as to selec- tion, and through experience he will learn to Primary Principles of Chartography 139 choose the method best adapted to each of the varying problems. importance of the right method The importance of selecting the best method is emphasized by Brinton in his Graphic Methods for Presenting Facts. He says: "After a person has collected data and studied a proposition with great care so that his own mind is made up as to the best solution for the problem, he is apt to feel that his work is about completed. Usually, however, when his own mind is made up, his task is only half done. The larger and more difficult part of the work is to convince the minds of others that the proposed solution is the best one — that all the recommendations are really necessary. Time after time it hapi)ens that some ignorant or presumptuous member of a committee or a board of directors will upset the carefully thought out plan of a man who knows the facts, simply because the man with the facts cannot present his facts readil}' enough to over- come the opposition. It is often with impotent exasperation that a person having the knowledge sees some fallacious conclusion accojited, or some wrong policy adopted, just becau.se known facts cannot be marshalled and presented in such a manner as to Ije effective. 140 Chartography in Ten Lessons "Millions of dollars yearly are spent in tb^ collection of data, with the fond expectatiofi that the data will automatically cause the con cc tion of the conditions studied. Though acrv data and real facts are valuable, when it getting results the manner of preseni ordinarily more important ti-an the factb . selves. The foundation of an edifice is of vasi importance. Still, it is not the foundation but the structure built upon the foundation who- gives the result for which the whole work \N'ae. planned. As the cathedral is to its foundation so is an effective presentation of facts to the data." ESSENTIALS OF GOOD CHART-MAKING The primary essentials of good chart-making are simplicity and clearness. The curves or bars of a chart are employed to express and to com- municate ideas, just as words are used in the English language. The fewer the ideas it is attempted to express in a single chart, the better. In fact, a single chart should aim to express only a single idea. This is diflBcult to accomplish, as the essence of a chart is a relation or a compari- son and this usually involves more than one idea. The aim, however, should be to construct the chart so that all but the dominant idea it is intended to express or communicate is kept Primary Principles of Chartography 141 subordinate or in the background. There should not be a single unnecessary mark or figure or word on the completed chart, and if its full meaning cannot be grasped quickly, then it has failed of its object. There is a common and quite general violation of ^hese principles. Chart-making is not at all 'omplex; it does not involve a knowledge of nigher mathematics for correct presentation and interpretation. There are a few definite rules which, if once understood, result in the ease and facility that may be likened to a knowledge of the alphabet, once acquired. Much of bad chart- making and of confusion in interpretation flows from a violation of the few simple princii)les. PLANNING THE SIZE OF THE CHART One task that will likely confront the student with as many peri)le.\ities as any other will be the working out of the size of the chart within the limitations of the sheet and the requirements of the statistics. Only practice will enable him in time to overcome most of these didiculties. Hut as a .sort of guide for meeting some of these prob- lems there is presented in the following paragraph a practical illustration. The size of the sheet for the completed chart is arbitrarily fixed for us at 12 by 16 inches. At 142 Chartography in Ten Lessons least two of the 16 inches are required as a margin on the left of the sheet (the chart appearing lengthwise) for binding. Usually an inch margin on the remaining three edges is advisable. This leaves 13 of the original 16 and 10 of the original 12 inches as the size within which the chartog- rapher is to work, or a size 10 by 13 inches. The neat lines of the frame and the letters of the title must come within this size. Usually one-half inch is sufficient for the title letters, and in our particu- lar illustration it is required that the title appear lengthwise of the sheet. This reduces the size to 9.5 by 13 inches. Between each of the neat lines and each of the scale lines one-half inch will usually be sufficient — this is a reduction in both dimensions of another inch. Sometimes a full inch is required below the lower horizontal line for the footnotes. Of course these spacings are subject to being increased or decreased according to the requirements of varying problems. The original size of 12 by 16 inches has dwindled by the above mentioned processes to 8.5 by 12 inches as the size of the framework proper. The arbitrary limitations of space within which the chartographer is confined in his work cannot be removed from among the difficulties of the practice of the art. All that he can do is to learn by experience to make the best adjustment pos- Primary Principles of Chartography 143 sible in each particular problem. This is true no matter what the size is that is determined upon. And having made the decision the chartographer will soon learn to adapt himself to the limitations of space and to forego something that is desirable in order to adjust his work to the exigencies of the requirements. In most cases where a number and variety of charts are to be filed for reference or bound as exhibits it is quite important that the size for all the sheets be made uniform. This does not neces- sarily mean that the worker will have the same size for the original of all the charts themselves, but it does mean that the completed charts shall all be on slieets of the same size. This permits of uniformity in size for all completed charts and assists in securing neatness and orderliness in office records. Completed charts on sheets of different sizes are awkward in handling and easily damaged. PLANNINO A REDUCTION IN SIZE Where the original <lrawing is to be reproduced by one of the .several j)hotograi)hic i)rocesses and printed from plates, the chartographer has a .special problem of reduction in size to .solve. In such casos the i)f'ri-aiid-ink charf should always be considoral)ly hirgor than the final r<'i)r<Mliiction. Most charts will stand a reduction in size of from 144 Chartography in Ten Lessons one-third to one-half and in cases even more, and will be improved in appearance by the process. A reduction in the size of the original drawing tends to smooth out the rough places or minor irregularities of lines, curves, bars, figures, and letters and results in a much cleaner effect. Virtually all the charts in these lessons have been reduced approximately one-half from the size of their original drawings. In ascertaining the dimensions for the size of the original drawing simply apply the rules of proportion. Only four factors are involved — the width and length of the reproduced chart and the width and length of the size that must be drawn to secure the reproduced size. Assuming our problem to be the one mentioned on page 142, we know the width and length of the reproduced size — the former is 10 inches and the latter 13 inches. We know also how much of a reduction we desire to secure — whether one- third or one-half and so on. Selecting a reduction of one-third gives us the arbitrary width of the original as 15 inches. It is the length of the original that must next be learned. Our known figures give us this formula: 10 : 13 :: 15 : X, which reads 10 is to 13 as is 15 to x. Working out this formula we learn that 13 times 15 equals 195 and this number divided by 10 gives Primary Principles of Chartography 145 19.5. This latter thus represents our unknown fourth quantity, which is the dimension of the length of the original drawing. The size of the original must then be 15 by 19.5 inches to secure a reduced size of 10 by 13 inches. In the above illustration the number 10 and the letter x are known as extremes and the numbers 13 and 15 as means. In any proportion the product of the extremes is equal to the product of the means, that is, 10 times x, the latter being 19.5, is 195, the product of the extremes, and 13 times 15 is 195, the product of the means. It is also true that the product of the extremes divided by either mean gives the other mean, as for instance: The product of the extremes is 195 (10 times 19.5) and 195 divided by 13, one of the means, gives 15, the other mean, or 195 divided by 15 gives IS. Again, the product of the means divided by either extreme gives the other extreme. For illustration: 13 times 15 is 195, the product of the means, and divided by 10 gives 19.5, or by 19.5 gives 10. Another simi)le method for ascertaining the reduced dimensions from the original is illustrated on the following i>age. The larger rectangle is our original she. From its lower left hand corner a diagonal line is drawn in light lead jjencil to the upper right-hand corner. This is line C-C. This Primary Prixciples of Chartography 147 diagonal line is then connected by a vertical line starting at any point on the base line A-A, such as the broken line shown. From the junction point of the broken line and the diagonal line draw another broken line at right angles to the vertical line and extending to line D-D. The rectangle thus cir- cumscribed in the lower left hand corner will be found to be in exact proportion to the larger rec- tangle, or the size of the original. In planning a reduction in the size of the chart from the original, care should be exercised in .seeing that all the lines on the original are made sufficiently wide to stand the reduction in line width due to the decrease in size. In our preced- ing illustration on j)age 144 the lines would be only one-tliird as wide in the completed as in the original drawing. Therefore, all the lines and curves and .so on of the original must be made wider than would be necessary if the chart were not to be reduced. Failure to allow for this reduc- tion in the width of the lines and curves and let- ters and figures is a common mistake made by the beginner in chartogrni)hy which should be guarded against if the best results are to be secured. TIfK UKDrriNfi c;la88 A valnable aid in this branch of tiic work is the reducing gla.ss. It nuiy l>e .said to be the opposite 148 Chartography in Ten Lessons of the magnifying glass, decreasing instead of increasing the size of the object observed, the lens being ground concave instead of convex. A convenient size is one with a single lens about one and three-fourths inches in diameter. This permits the lines, figures, and letters on a chart to be seen in sizes from one-half to one-fourth smaller than their originals. In observing the parts of the chart for an in- dication of the size of the proposed reduction the most accurate method is to hold the glass at different distances above the sheet so that looking through it with the left eye two or three or four squares, depending upon the amount of reduction desired, equals one square as seen by the unob- structed right eye. Thus by superimposing and comparing the images of both eyes the required reduction can be measured. This enables a comparison of the width of lines, figures, and letters as originally placed with their width after reduction and permits the determination as to whether they must be made still wider or heavier or larger in order to meet the reduced size. Even with the use of the reducing glass the be- ginner is likely to find at first that his lines, curves, figures, letters, and so on when reproduced do not appear to be as heavy or as large as he had anticipated. Primary Principles of Chartography 149 Without the emploj'ment of these rules and aids in reduction one has to depend largely upon guess work as to whether the chart and its parts will present a proper appearance when reduced, and guess work, as has been said, should be eliminated from chartography. The student must not forget that accuracy is a valuable mental quality which is useful elsewhere than in chartography', and if the practice of this art teaches it to him he has gained an additional asset of great usefulness. THE FRENCH CURVES Another tool that the student may find useful, especially in cliarting curves, is what is known as the "PVench curves." These are on sale at any first-class store dealing in drafting instruments. While they do not always in their entirety fit into the complete curve the student luis to make, they can be shifted forward or backward so as to cover fairly accurately the lead-j)encil dots measuring the points of contact of flie curve. Usually they give a clean, smootli curve if care is exercised in their use. In most charts where the scale niiils permit the curve to move regularly uj) or down across the sheet, the curve appears smooth without sharp movements that result in peaks. In all the 150 Chartography in Ten Lessons illustrations in these Lessons these curves have been drawn in freehand, and this method is recommended as satisfactory. This peak-top does not indicate as minute a degree of accuracy in the figures upon which the curve is based as does the smooth curve. OTHER MECHANICAL AIDS One serious drawback in making the letters by hand is the length of time required, even after one becomes proficient in lettering. Under conditions where the number of charts to be drawn is large, efficiency is best served and the cost of produc- tion materially reduced if recourse is had to a small printing press with about three fonts of type of 18, 12, and 10 point, commercial Gothic. The moderate expenditure will soon be com- pensated by using for other work the time saved from lettering by hand. Printed letters photo- graph satisfactorily in almost any process of reproduction. Another recourse instead of drawing letters by hand is to make use of gummed black paper letters and figures which are for sale at first- class stationery stores. These can usually be pasted quite neatly on the chart that is to be reproduced, if a light pencil line is made for a guide along the bottom of the spacing for the Primary Principles of Chartography 151 letters. This pencil line must of course afterwards be erased. CHECKING-UP THE COMPLETED CHART After the drawing has been finished there remains for the student a very important task. The explanation of this task in detail involves describing the concrete application of all the rules and principles of chartography that have been observed in the construction of the chart. This is true in the .sense that the student must see to it, by a rigid and thorough checking-up of the lines and figures and letters and so on before the chart leaves his hands, that all these rules have been strictly applied. As related to the check- ing up of certain features of the curve chart, this ta.sk ha.s already been referred to in I^esson IV, pages 50 and 51. The immediately follow- ing statements apply particularly to the bar chart. Each bar should extend to the point on the chart that its statistical number as measured by ^ the scale determines — it should neither fall short of this i)oint nor extend beyond it. Be sure that each bar is f)roperly spaced from adjoining bars. In making the bars of a chart it is quite often possible that all the space to the right of the vertical scale or column of years and to the left of 152 Chartography in Ten Lessons the ends of the bars, and from the horizontal scale line to the base line, can be made black with a small brush dipped in India ink, the ends of the bars being squared with the pen. After the ink dries the bars can easily be outlined and separated from each other according to the ver- tical scale units by drawing horizontal lines in Chinese white. This expedient enables a great deal of work to be done in a comparatively short space of time, and the results are highly satis- factory. When the India ink is used in this way it is advisable to apply two or more coats or washes in order to insure a uniform density of surface. Chinese white is an opaque composition which may be thinned down to desired consistency by the addition of water. Besides its use in separating bars out of a block of black, Chinese white is also excellent for concealing black ink lines or marks where erasure is impracticable. The scale units should be correct at each of the points of measurement and neither to the right nor left of their proper places. The chart should imclude the statistical table from which it has been made. If it is found impossible to include the statistical table on the chart it should be on an accompany- ing or attached slip or sheet. Primary Principles of Chartography 153 The table of figures should be correct and neat and properly located and boxed without crowding. The period of time column both in the table and adjacent to the bars must be correctly and properly alligned. Be sure that no mistake has been made in copy- ing any figures on to the chart. The statistical table involving periods of time should be presented with the earliest period first. The source of the statistical table should in every instance be given, preferably in the foot- notes. Instead of checking up the movement of the bars or curves from either the statistical table on the chart or the scale figures tl)emselves, reference should be had to the original figures. All additions, subtractions, multiplications, divisions, and so on derived from the statistical table should be computed at least twice and by different f>ersons. See that all horizontal lettering reads from left ^ to right and all vertical lettering from the base of the chart uj>ward. The title should l>e clear and concise and yet comprehensive. It should have every word spelled correctly and shouhl contain the fewest possible words consistent with clearness of ex- pression. It sometimes happens that the title 154 Chartography in Ten Lessons can be improved in these directions over the first selection of words after the student has been working with the statistical material. Words as well as letters in the title should be evenly placed and spaced — none of them must be askew. THE PROCEDURE IN CHECKING UP It is an advantage in checking up a chart to start with the title, next the horizontal scale, then the vertical scale, the table of statistics, and then relate the movement of the curves or bars to these factors. The horizontal and vertical lines forming the background of the chart must not be overlooked as to their proper distance apart. The foot-notes and the neat lines require equally careful attention. Make sure that the neat lines are wider or heavier than the vertical and horizontal lines of the framework. Foot-notes should be as brief as possible con- sistent with clearness, should read from left to right, and should not be askew. In selecting the designations for the curves and bars, have the most conspicuous been made to correspond with the particular statistical element it is desired to emphasize? For instance, in bar designations the solid black is generally more noticeable. Primary Principles of Chartography 155 Check carefully the key or legend designations with those of the curves and bars to see that they correspond accurately. Have the proportions been correctlj^ determined for the required reduction? Be especially careful that all lead pencil and unnecessary ink marks, used as guides or other- wise, have been erased or removed. If not, in case of reproduction these are likely to photograph and thus affect disadvantageously the neat ap- pearance of the chart. CLEANLINESS ESSENTIAL TO NEATNESS In handling a chart keep the hands clean, especially from the drawing ink which will smear the sheet. .Vn aid to this will be found by keeping the ink bottle on a })lotter which will absorb drops and i)revcnt them from getting on the drawing board or table. Blot immediately every ink spot. Never fold a rhart. Keep the .sheet flat or roll it. A folded chart cracks or crea.ses the sheet and breaks the lines, bars, curves, and .so on. All checking and verification should be done also by some f)iic other than the |>orson who drew the chart so that there may \)v greater certainly in the detection and correction of errors. For even the best chartograjjher nuikes mistakes. The student should assure himself before per- 156 Chartography in Ten Lessons mitting the chart to leave his hands as completed that in all respects it is in condition to receive his final O. K. PLOTTING THE CHART IN ROUGH OUTLINE After the chartograplier has completed his checking up he should devote several minutes to a consideration of the possibility that he might have selected a different method which would have brought out the point of the statistics more clearly. He will have fewer regrets if he adopts and consistently follows the practice of first plotting his chart in rough outline in lead pencil at the very outset of his work. He should apply this to several methods before finally determining upon any. He will find that though this practice takes a little time at first it will in the end greatly expedite his work. It should never be forgotten that as chartog- raphy primarily supplements statistics with the object of making them clear and comprehensible at a glance, a chart that is not more clear in exposition than the statistical data upon which it is based has missed its object. Also it should be remembered that the chart " tells the story " — it should need very little explanation, if any. Instructions to the lithographer should be clear and definite. These may be written on a slip of Primary Prixciples of Chartography 157 paper and attached to the drawing with a dip, but a safer plan is to write them on the coordinate sheet itself, preferably on the back. Before sending the drawing to be reproduced be sure to cut away the margin of the sheet where the thumb tacks have held it in place on the drawing board, as these punctures are likely to show in the reproduced chart. Even when the drawing is not to be photographed, such punc- tures in the paper detract from the neat ap- pearance of the finished chart. Before deciding upon the uniform size of .sheet for a number of different kinds of charts it is advisa})le to consult the lithographer in order that a size may be selected which permits of the least possible waste or loss in cutting from the larger sheet. It is imf)()rtant also to examine closely into the quality of the {lajjcr of the reproduced chart. The preservation of the chart depends to a large degree upon this quality. Paper containing sulphite pulj) or other chemicals suffers rapid deterioration. U'ithiri a short time such paper becomes brittle and di.scolored, and these defects seriously affect the preservatif)ti of chart records for any length of time. A high grade linen bond paper, although it costs more i)er sheet, is less expensive in the long run. The checking up of the completed chart as well 158 Chartography in Ten Lessons as the details of the planning and drawing should impress upon the student the necessity for observ- ing closely and carefully every factor with which he deals. He cannot afford to overlook even the smallest detail as every detail must be accurate if the completed chart is to be accurate. This attention to minor details fixes the mind upon the correctness of the figures; on the accuracy of the scale units; on the spacing of figures and lines; on the spelling of words; on the correct designation and spacing and length of the bars, and so on. Looking for possible defects or errors develops the critical faculties. In the course of practice all these separate but important details soon fix a habit of mind and that which at first may be hard work sooner or later becomes almost mechanical attention. Working with tools that require accuracy in their use the chartographer soon learns from mistakes he must correct that it does not pay to make mistakes — that it is a loss of time and energy and materials — and he comes consciously to apply himself so as to avoid their repetition in order not to be compelled to do the work over again. Out of this experience he learns eflSciency in the concen- tration of his energies and in their application to his specific task. Chartography is thus an invaluable mental training. It lends accuracy to Primary Principles of Chartography 159 constructive thinking; it leads to the further study , of statistics and of the underlying forces back of them, and by these and similar steps the pro- gressive student develops his thinking caj)abilities. QUESTIONS FOR SELF-EXAMINATION 1. What is meant by planning the chart? How does it differ from plotting the chart? 2. Diacuss the importance of selecting the right method. 5. What are the essentials of good chartography? 4. How is the size of the chart planned? 5 How is a reduction in size of the original determined? 6. Of what assistance is the reducing glass? 7. What are "French curves"? 8. Describe the more important phases of checking up the completed chart. 9. Summarize briefly the more important of the funda- mental principles of chartography. 10. Of what value is chartography in training the mind? This book is DUE on the Rist date stamped below i^'-f JAN 4 1937,^ipf^ Decioi95f DEC i ORECO •m L-9-15»i-7,'32 JAIM ii ■' SEP 11961 ^ fW74 Wy 3 7/975 UC SOUTHERN REGIONAL LIBRARY FACILITY ||!„|lllllllll>lll1ll AA' 000 561 648 7