r Z 2S SCHOOL — iPtTBLICATIOJNTS OJ^THE Clark University Library WORCESTER, MASS. Editkd by JLOtTIS N. WILSON, Litt. D., LroBAHiAN Vol. 3. January, 1912. No. 1. The Relative Legibility of Different Faces of Printing Types By BARBARA ELIZABETH ROETHLEIN, A. M. Clark University WORCESTER. MASS. f _ • • THE RELATIVE LEGIBILITY OF DIFFERENT FACES OF PRINTING TYPES By Barbara Elizabeth Roethlein, A.M., Clark University Communicated by John Wallace Baird I. Introductory I II. Historical 2 III. Experimental 5 a. Materials and Apparatus 5 b. Method of Procedure ...... 7 c. Observers ........ 8 d. Results 9 IV. Discussion of Results ....... 10 V. Conclusions 33 VI. Bibliography 34 I. INTRODUCTORY The aim of the investigation which is here reported was to determine the relative ease or difficulty with which various 'faces'^ or forms of printed letters can be read; and to discover what relationship obtains between legibility and cer- tain definite modifications of 'face.' A number of typically different faces of type were selected for investigation; and our experimental procedure consisted essentially in determin- ing to what extent the legibility of each face was affected — both when the letters were presented in isolation and in groups — by the introduction of unfavorable conditions for reading. The present paper will deal only with 'type-faces' ; the question of the part which is played by printing papers and printing inks will be discussed in a later paper. ^ The term ' face ' is employed here and throughout this paper in the sense in which it has come to be used by printers and type- founders. It is customary to group the numerous variants of letter- form into families, — Caslon, Cheltenham, Jenson, Ronaldson, etc., — and to speak of each family or typical variant from the common letter- form as a ' face.' 247891 2 ROETHLEIN At the time when the art of printing from individual or moveable types was first introduced, the forms of the letters of the alphabet were few in number and exceedingly crude in design. The use of 'black letter' had been popularized by generations of manuscript writers ; and the early printers were content to appropriate those letter-forms which they found to be in current use. But within a few decades designers and draughtsmen set themselves the task of simpli- fying and improving the existing forms of letters ; and, indeed, it seems probable that the Roman 'faces' were introduced about the year 1465, and the Italic 'faces' some thirty-five years later. The ingenuity of many generations of mediaeval and modem designers has produced a multitude of variants of letter- forms, many of which are familiar to every reader (a, a, a, a, a; g, g, g, g, g). Hundreds of different 'faces' of type have been designed and ptit upon the market; and while it is true that certain of these 'faces' are employed only for purposes of ornamentation and display, yet an enor- mous variety of letter-forms is to be found in our books and magazines. Every reader has observed that all of these variants of letter-form are not equally legible — an obser- vation which raises the theoretical question: What are the factors upon which legibility depends? And the practical question : How should one proceed if one set out to improve the legibility of printed letters? II. HISTORICAL Psychologists have been engaged these many years in an investigation of the act of reading in its various aspects. But there is a singular dearth in the literature so far as the specific topic of the present investigation is concerned. More than forty years ago, Exner (17) and Baxt (2) undertook to measure the brief period of time which is necessary for the perception of visual objects (letters, words) ; and in 1885 Cattell (6, 7, 8) continued the investigation of the same problem. Baxt had reported that, under his most favorable conditions of illumination, it was possible to read two or three letters of a total group of seven when LEGIBILITY OF DIFFERENT FACES OF PRINTING TYPES 3 they were exposed for one one-hundredth of a second. Cattell devised a falHng-screen apparatus which enabled him to vary and to measure his exposure-times. He found that the differences in the times which are necessary for the recognition of isolated letters, whether upper case or lower case,^ whether Latin or German, are of negligible magnitude. When the exposure-time was very brief, it was found that the letters were not always read correctly. A record of the percentage of correct readings of the various letters, when presented under constant and uniform conditions, enabled Cattell to determine the relative legibility of the letters. The order of legibility (descending) was found to be: WZMDHKNXAYOGLQISCTRPBVFUJE and dkmqhbpwuljtvzrofnaxyeigc s, — W being read correctly in 89 per cent, of the trials, E in 23 per cent. ; d in 87, and s in 28 per cent. Sanford (35) employed a similar method, but obtained a somewhat different result; his order was (for Snellen type, lower case) : mwdqvyjp kfblighgrxt ouanescz Sanford also determined the relative legibility of the same letters by a distance method, and obtained the following result : wmqpvyjf hrdgkbxlnu atizocse Sanford also tested alphabets representing two other letter- forms, — a modern face, and a bold oldstyle face. The modern face letters were recognized in the following order (distance test) : dpqmyknw ogvxhbjlia tuzrscfe The oldstyle letters fell into the following order (tested by the method of brief exposure) : mwpqvykb djrlonighu atfsxzce Finzi (18) employed the method of brief exposure, pre- senting a group of nine letters at each exposure. From ' The term ' upper case ' will be used throughout to designate the capital letters, and ' lower case ' to designate the small letters. 4 ROETHLEIN the percentage of misreadings of each letter he computed the order of legibility as follows : P U A Q X T D S E W AI V Y Z H C N F L R G B K O I, the percentage of errors ranging between .8 for P, and 7.8 for I. The particular type or letter-form which Finzi em- ployed is not specified in his paper. Griffing and Franz (21) investigated the influence of size and form of letter upon legibility. Their experiments com- prised a fourfold test : What is the difference in the rapidity with which small print and large print can be read (five- point and tw^elve-point, both Roman) ? How many letters of each of these two sizes can be read in a single brief exp: sure? How long must letters of each size be exposed in order to insure their correct reading? What intensity of illumination is necessary for the recognition of letters of various forms and sizes, — Roman letters, .8 mm. and 1.6 mm. high; Gothic letters, .9, 1.6, 3.1, and 6.0 mm. high? These investigators found that their larger types were, in every instance, more legible than their sm.aller types ; and that Gothic letters were m.ore legible than Roman letters. Besides these experiments which have just been described, numerous attempts have been made to investigate other factors which have to do with the act of reading. The nature and the extent of the eye-movements by means of which the reader follow^s the printed line have been examined and measured by Huey (22, 23, 24), Dodge (13, 14, 15), Erdmann and Dodge (16), Dearborn (10), and others. It has been estab- lished that the movement of the reader's eyes does not pro- ceed gradually and continuously across the page, in any such fashion as, for example, a meteor moves across the sky. Typical eye-movements consist of a succession of alternate leaps and pauses ; nor is the movement always in a forward direction, because it frequently happens that one's eye-movement proceeds backward, i.e., to the left, from an intermediate fixation-point. The number of pauses may vary from two to seven in a line whose length is twenty centimeters ; but the usual distance between successive fix- ation points or pauses is approximately 2 cm. It seems LEGIBILITY OF DIFFERENT FACES OF PRINTING TYPES $ probable that no words are seen while the eyes are in move- ment; and that the act of reading a printed line consists of a series of interrupted glimpses, during each of which one reads a small section which extends to the right and to the left of the fixation-point. This discovery makes it seem probable that a considerable part of the printed line is imaged, not upon the fovea, but upon para-foveal regions of the retina. And the capacity of these paracentral regions to distinguish the forms of letters be- comes an important topic for investigation. This problem has, indeed, been attacked by Kirschmann (28) and by Dockeray (12) ; but further investigation is needed before one can make any definite statement regarding the legibility of letters in indirect vision. Numerous other investigations of the problems of reading have been made by Babbage (i), Becher (4), Goldscheider and Miiller (19), Javal (26, 27), Maire (31), Messmer (32), Pillsbury (33), Quantz (34), Schumann (t,/), and Zeitler (40) ; but it seems more appropriate to discuss their results in connection with our own findings, than to summarize them in this section. III. EXPERIMENTAL A. ISOLATED LETTERS a. Materials and Apparatus The materials which were employed in the present in- vestigation were sheets of printed letters ; and the apparatus consisted of a mechanism by means of which these letters could be presented at a variable distance from the observer. When the investigation was first undertaken, we made a careful examination of the various faces of type which are listed in the sample-books of the American Type Founders Company.^ We selected fifty faces of type, — comprising ' The author is indebted to the American Typefounders Company, Jersey City, for a liberal donation of types and of prints, without which the investigation would have been impossible. Especial thanks are due to Messrs. Frank B. Berry, L. B. Benton and Morris Benton of that firm for valuable suggestions regarding letter-forms and 6 ROETHLEIN some thirty ordinary faces, together with such variants as italic, bold, condensed, expanded and various combinations of these variants. In these earlier experiments, it was decided to employ the method of brief exposure; and an apparatus was devised which provided for a succession of exposures, each one one- thousandth of a second in duration. The series of exposures of any given letter was terminated by the depression of a key by the observer; and an automatic counting device re- corded the number of exposures which had been necessary for the reading of the letter. For certain technical reasons this apparatus and mode of procedure were abandoned after a number of preliminary experiments had been made; and the distance test was substituted.* The apparatus, by means of which we obtained the results on which this paper is based, consisted of a long bench along which moved a sliding carriage containing the letters to be read. The bench was 440 cm. long and 15 cm. wide. Its proximal end was 78 cm., and its distal end 63 cm. above the floor, so that the observer was able to assume the primary position of regard throughout. The higher end of the bench was provided with a vertical support, which carried a head rest, — the hood of a stereoscope. This device enabled the observer to assume and to maintain a constant position in relation to any point on the scale upon the side of the bench. regarding the interpretation of our results. We are also under obli- gation to Dr. H. L. Koopman, of Brown University, and to Mr. C. Chester Lane of the Harvard Press; Mr. L. D. Evans of the River- side Press, Cambridge, Mass., has furnished us with valuable data concerning compositors' errors and proof-readers' errors. * It is a well-known fact that concentration of attention has a very pronounced efifect upon reaction-time, and that the duration of the reaction varies with variations in degree of concentration. It seemed difficult, even impossible, to maintain the same degree of concen- tration through thousands of readings of letters ; and, in the absence of a control of this exceedingly influential factor, the results of our tachistoscopic experiments seemed wholly unreliable. For this reason the tachistoscope was abandoned, and a method which consisted in pre- senting the letters at variable distances was substituted. LEGIBILITY OF DIFFERENT FACES OF PRINTING TYPES 7 The carriage which moved along the bench consisted of a box 55 cm. high, 40 cm. wide, and 25 cm. deep. The front of the box had been cut away ; and the back consisted of a wooden wall against which the sheet of printed letters was attached and held in place by a sheet of glass. The back wall of the box was illuminated by a number of electric lamps, so arranged that the whole surface of the sheet of letters was uniformly illuminated. To the side of the carriage was attached an indicator, which just cleared a metric scale upon the side of the bench ; this device enabled the experimenter to ascertain the distance of the sheet of letters from the eye of the observer at any given setting of the carriage. The sheets upon which the isolated letters were printed were 21 cm. wide and 36 cm. long; and the paper of all of the sheets was of the same quality and texture.^ Each sheet contained twenty-eight letters, all of the same face and of the same case, — the complete alphabet, with two of its letters repeated. The letters were arranged in random sequence, in four lines ; they were so spaced that each letter stood at a distance of 3.7 cm. from its nearest neighbors on the same and on adjacent lines. Fifty-two different sheets of letters were investigated, representing the following twenty-six faces of type, both lower case and upper case. All of our letters were of the size which is technically described as ten-point ; the reader will find them illustrated in Tables I and II. (See inserts between pages 8 and 9.) American Typewriter Bold Antique Bulfinch Caslon Oldstyle No. 540 Century Oldstyle Century Oldstyle, Bold Century Expanded Cheltenham Oldstyle " This paper is technically described by the manufacturers as a white, coated book-paper, 25 x 30 — 80. a ROETHLEIN Cheltenham Bold Cheltenham Bold, Condensed Cheltenham Italic Cheltenham Wide Clearface Clearface Bold Clearface Italic Clearface Bold Italic Cushing No. 2 Cushing Oldstyle No. 2 Cushing Monotone Delia Robbia DeVinne No. 2 DeVinne No. 2, Italic Franklin Gothic Jenson Oldstyle No. 2 News Gothic Ronaldson Oldstyle No. 551 b. Method of Procedure All of the readings were made in a semi-darkened room, in order that the (artificial) illumination upon the sheet of letters might be controlled and kept constant throughout. The experimental procedure was as follows : After the observer had become adapted to the illumination of the room, a sheet of letters was placed in position in the carriage and the series of readings began with the carriage at the farther end of the bench. The observer had been instructed to read the letters at a uniform tempo, substituting "blank" for the name of any letter which was not easily decipherable. This precaution seemed necessary to prevent the observer from giving an undue amount of attention to any one letter of the series at the expense of the other letters, i. e., to prevent him from puzzling longer over one member of the series than over any other member. After he had thus attempted to read through the complete list of letters, the carriage was moved to a point twenty centimeters nearer his eye than the initial setting; and he made a second attempt to read LEGIBILITY OF DIFFERENT FACES OF PRINTING TYPES 9 the letters. In the first and alternate readings of each series he began at the upper left-hand corner of the sheet and proceeded from left to right along each line, taking the lines, in order, from the top downward. In the second and alter- nate readings, he began at the lower right-hand corner of the sheet and proceeded from right to left, and from below upwards. This procedure was continued, the carriage being advanced step by step, until every letter upon the sheet had been identified. The experimenter was provided with a duplicate sheet of letters, similar to the one which had been inserted in the carriage ; and upon this duplicate sheet she recorded the misreadings, and the farthest distance at which each letter was read. Each of the fifty-two sheets of letters was read twice by each of the six observers. Before being inserted in the carriage for its second reading, each sheet was cut along its longitudinal and its transverse diameters ; and the four quarter-sheets v/ere reassembled in such fashion that those letters which had formerly appeared upon the marginal regions of the original sheet now appeared upon central regions of the reconstructed sheet, and vice versa. A period of several days always elapsed between the first and the second reading of any sheet. c. Observers The observers were instructors or students in the depart- ment: Messrs. R. Acher, J. W. Baird, E. O. Finkenbinder, F. A. Lombard, H. B. Moyle, and C. W. St. John ; they all possessed emmetropic or adequately corrected vision. Each observer gave a complete series of one hundred and four readings. B. GROUPED LETTERS In the second group of experiments the apparatus and the method remained unchanged, but here the letters were presented in groups instead of singly. In these later experi- ments only lower case letters were employed. Eight faces were elected from the twenty-six which had already been lO ROETHLEIN used, and Scotch Roman was added, — the complete list of nine faces being as follows : Bulfinch Caslon Oldstyle No. 540 Century Oldstyle Century Expanded Cheltenham Wide Cushing- Oldstyle No. 2 Cushing A-Ionotone News Gothic Scotch Roman Each group of letters formed a nonsense combination; and the groups of each face were arranged in three lines upon sheets of the same size as had been employed in the former experiments. Sets of grouped letters were printed, in duplicate, upon coated book paper, of the good quality and upon an antique laid book paper of the same quality and weight as the paper of this page. Our mode of grouping the letters aimed to introduce as many difficult and confusing combinations of letters as pos- sible. We were guided in the combining of the letters into groups by data furnished by Mr. L. D. Evans, an expert proofreader, and by the confusions which had been recorded in our earlier experiments. The combinations of letters which are here appended illustrate a typical series of groups ; they also illustrate the Scotch Roman face. ksitugy cdzxpbj ftoceygqa wrvlindh hknurfkxzqg munimm bhwvjyst oceo wvxarlizxp ybhdonactilsf dnupqcetrlj Only three observers, Messrs. Baird, Finkenbinder and St. John, took part in these experiments. Each observer gave two readings of each face, — an average of twenty-four readings of each letter of each face. TABLE I. UPPER CASE. ISOLATED LETTERS Showing the Average Distance, Expressed in Cm., at Which Each Letter op Each Face was Read (Twelve Readings, Six Observers). The First Column in Each Division of the Table Shows the Actual Size and Form op the Letter Which was Presented for Identification; and the Number Indicates the Average Distance, from the Eye, at Which the Letter was Correctly Identified American Typewriter Bulfinch Caslon 0. S. Century 0. S. Century Expanded Cheltenham O.S. Cheltenham Wide Clearface dishing 0. S. Cushing No. 2 Cushing Monotone Delia Robbia De Vinne No. 2 Jenson 0. S. News Gothic Roualdson O.S. Average A 221.7 A 280 A 270 A 300 A 300 A 291.7 A 281.9 A 257.5 A 241.7 A 234.2 A 293.3 A 265 A 300 A 286.7 A 263.3 272 A B 176.7 B 231.7 B 206.7 B 221.7 B 213.3 B 210 B 221.7 B 204.2 B 190.8 B 179.2 B 211.7 B 211.7 B 229.2 B 230.8 B 195 208.9 C 195 c 283.3 C 278.3 C 285 C 271.2 C 290 C 288.8 C 259.1 C 243.3 C 251.7 C 282. 5 C 252.5 C 310 C 276.7 C 259.2 265.1 D 1S3.3 D 275 D 265 D 281.7 D 273.3 D 270 D 251.7 D 255.8 D 218.3 D 236.7 D 243.3 D 249.2 D 300 D 265 D 246.7 254.3 E 193.3 E. 230.8 E 213.3 E 241.7 E 240.8 E 259.2 s^H E 229.6 E 198.3 E 209.2 E 192.5 E 228.3 E 226.7 E 245. S E 240 E 208.3 223.9 F 191.7 F 254.2 F 210.8 F 265 F 265.8 F 241.7 I5 F 270.7 F 234.2 F 208.3 F 217.5 F 244.2 F 243.3 F 276.7 F 253.3 F 241.7 241.6 G 169.2 G 276.7 G 256.7 G 265 G 255 G 275 0^ G 258.6 G 220 G 197.5 G 214.2 G 256.7 Q 1S4.2 G 273.3 G 247.5 G 213.3 237.6 H 190 H 266.7 H 230 H 250 H 239.2 H 256.7 H 261.2 H 213.3 H 202.5 H 214.2 H 283.3 H 233.3 H 273.3 H 253.3 H 240 240.5 I 230 1 300 I 285 I 295 I 275 I 268.3 I 327.2 I 290 1 244.2 I 251.7 1 301.7 I 273.3 I 296.7 1 296.7 I 270 280.4 J 226.7 J 290 J 285 J 300 J 301.7 J [281.7 P J 323 J 295 J 271.7 J 260 J 315 J 275.0 J 320 J 295 J 281.7 287.5 K 177.5 K 262.5 K 220.8 K 256,7 K 241.7 K 254.2 S 5- K 236 K 228.3 K 205.8 K 200 K 217.5 K 243.3 K 270 K 231.7 K 229.2 231.7 L 248.3 L 305 L 281.7 L 308.3 L 301.7 L 300 sa L 315.1 L 296.7 L 261.7 L 266.7 L 296.7 L 290 L 311.7 L 300 L 283.3 291.1 M 189.2 M 316.7 M 290 M 313,3 M 298.3 M 303.3 Oo M 323 M 315 M 273,3 M 275 M 321.8 M 293.3 M 318,3 M 308.3 M 268.3 293.8 N 184.2 N 246.7 N 244.2 N 258.3 N 241.7 N 261.7 r^^O N 247 N 221.7 N 201.7 N 215 N 258.3 N 225 N 270 N 243.3 N 225 235.5 186.7 O 293.3 O 267.5 293.3 265 O 278.3 274.5 230.8 215.8 219.2 275 226.7 296.7 250 253.3 254.0 P 190 P 301.7 P 245 P 263.3 P 281.7 P 270.8 ^ c?" P 286 P 255 P 238.3 P 229.2 P 252.5 P 263.3 P 296.7 P 278.3 P 229.2 257.9 ft 214.2 Q 308 . 3 Q 270 Q 268,3 Q 275 Q 291.7 Q 285 Q 255.8 Q 220 Q 245 Q 246.7 Q 240.9 Q 295 Q 278.3 Q 231.7 261.7 R 186.7 R 227.5 R 200.8 R 210.7 R 235 R 255 ~ R 216.9 R 198.3 R 196.7 R 180 R 228.3 R 193.3 R 234.2 p 240 R 205.8 214.0 S 168.3 5 236.7 S 197.5 S 225 S 209.2 S 215 ^ S 211.2 S 210 S 182.5 S 183.3 S 203.3 S 200 S 223.3 s 216.7 S 203.3 205.7 T 220.8 T 276.7 T 280 T 305 T 280 T 271.2 I' T 273.9 T 280 T 219.2 T 237.5 T 295 T 251.7 T 308.3 T 280.8 T 251.7 268.9 U 200.8 U 268.3 U 247.5 U 256 7 U 276.7 U 266.7 3 U 269.1 U 233.3 U 220 U 232.5 U 275 U 240.9 U 270.8 U 273.3 U 237.5 251.3 V 205 V 275 V 255.8 V 281.7 V 278.3 V 2SS.3 V 277.6 V 246.7 V 237.5 V 236.7 V 291.7 V 248.3 V 310 V 264.2 V 256.7 263.5 w 165 w 310.7 W 291.7 w 306.7 w 316.7 W 306.7 «) w 318.2 W 331.7 W 288.3 W 285 w 343,3 w 315 W 306.7 w 316.7 w 301.7 300.2 X 192.0 X 260.8 X 237.5 X 257.5 X 250 X 263.3 ?• X 247 X 240 X 212.5 X 222.6 X 252.5 X 236.7 X 256.7 X 240 X 228.3 239.8 y 221.7 Y 268.3 Y 260 Y 271.7 Y 261.7 Y 249.2 c_ Y 266 Y 250 Y 241.7 Y 233.3 Y 263.3 Y 225 Y 261.7 Y 258.3 Y 250.8 252.1 z 187.5 z 261.7 Z 228.3 Z 250 Z 237.5 Z 261.7 Z 240.7 Z 218.3 Z 204.2 Z 207.5 z 256.7 Z 216.7 Z 270 Z 255.8 Z 210.7 233.8 Average IQQ. 8 273.8 250.7 270.4 264.8 268.5 269.3 247.6 224.8 228.4 266.8 243,2 281.7 264.6 241.7 252.8 LEGIBILITY OF DIFFERENT FACES OF PRINTING TYPES II d. Results A. ISOLATED LETTERS The experimenter's record-sheet contained a statement of the distance at which each letter was read, together with a list of the confusions or misreadings, and an introspective description of the procedure which the observer had followed in deciphering the letters. The numerical data have been compiled and tabulated in various ways. Tables I and II show the averages of the numerical results, arranged in order of faces. These two tables report the data for 'ordinary' faces only, — the italic, the bold and the condensed faces not being included here. The numbers which appear in these tables indicate the averages of the extreme distances at which the letters were read, — hence the larger the number appended to any letter, the greater the legibility of that letter. Thus, in the first column of Table I, "H 190, I 230" may be taken to signify that the upper case I of the American Typewriter face is considerably more legible than the upper case H of the same face. Each vertical column of these two tables contains, there- fore, a statement of our findings regarding the relative legi- bility of the various letters of a given face; and the number at the foot of the column indicates the average legibility of the twenty-six letters of that face. The numbers in each horizontal line of the tables indicate the relative legibility of each of the sixteen variants of each of the twenty-six letter forms. Table III presents the grand averages of the sixteen faces; it also contains similar data for the bold and for the italic faces and for two extra-bold faces, Franklin Gothic and Bold Antique ; while Table IV shows the grand average distance at which each letter of the alphabet was read, the data being here compiled from the readings of the complete set of sixteen faces. The efifect of certain definite modifications of a given letter- form is shown in Tables V and VI. These two tables are com- piled from results which were obtained with Cheltenham Old- LEGIBILITY OF DIFFERENT FACES OF PRINTING TYPES II d. Results A. ISOLATED LETTERS The experimenter's record-sheet contained a statement of the distance at which each letter was read, together with a list of the confusions or misreadings, and an introspective description of the procedure which the observer had followed in deciphering the letters. The numerical data have been compiled and tabulated in various ways. Tables I and II show the averages of the numerical results, arranged in order of faces. These two tables report the data for 'ordinary' faces only, — the italic, the bold and the condensed faces not being included here. The numbers which appear in these tables indicate the averages of the extreme distances at which the letters were read, — hence the larger the number appended to any letter, the greater the legibility of that letter. Thus, in the first column of Table I, "H 190, I 230" may be taken to signify that the upper case I of the American Typewriter face is considerably more legible than the upper case H of the same face. Each vertical column of these two tables contains, there- fore, a statement of our findings regarding the relative legi- bility of the various letters of a given face; and the number at the foot of the column indicates the average legibility of the twenty-six letters of that face. The numbers in each horizontal line of the tables indicate the relative legibility of each of the sixteen variants of esch of the twenty-six letter forms. Table III presents the grand averages of the sixteen faces ; it also contains similar data for the bold and for the italic faces and for two extra-bold faces, Franklin Gothic and Bold Antique ; while Table IV shows the grand average distance at which each letter of the alphabet was read, the data being here compiled from the readings of the complete set of sixteen faces. The effect of certain definite modifications of a given letter- form is shown in Tables V and VI. These two tables are com- piled from results which were obtained with Cheltenham Old- 12 ROETHLEIN style presented in ordinary, in bold, in bold-condensed, (in wide,) and in italicized form. Tables VII, VIII, IX, and X show the five variants of each letter which proved to be most legible, and the five which proved to be least legible. B. GROUPED LETTERS The results of our second series of experiments are pre- sented in Tables XI and XII. Table XI shows the average distance at which each letter of each of the nine faces was read when presented in groups. The reader is warned against drawing any conclusion from this table regarding the relative legibility of the various letters of any given face, e.g., regarding the relative legibility of the Caslon m and the Caslon, or any other, k. In other words, the numbers v/hich appear in the horizontal lines are comparable with one an- other; but the numbers which appear in the vertical columns are incomparable with one another. Table XII presents the same results as are contained in Table XI, but they are now arranged in order of magnitude in order to show the (descending) order of legibility of the several variants of each letter-form. IV. DISCUSSION OF RESULTS A. ISOLATED LETTERS The problem with which we are here concerned may be given the following general formulation : Which of several geometrical figures is most clearly perceptible, and most readily distinguishable from other geometrical figures? But while the twenty-six letter-forms which constitute the alpha- bet may be regarded as a series of geometrical figures of dififerent shapes and of different degrees of complexity of detail, and while any twenty-five of the different twenty-six different faces of type which were employed in the present investigation may be regarded as variants from the twenty- sixth or common letter-form, yet, as a matter of fact, such a simple mode of envisagement of our problem does not do justice to the complex conditions which are found to be N IN Each Division op the Table Shows the Actual Size and Form of the ICH THE Letter was Correctly Identified a Robbia De Vinne No. 2 Jenson 0. S. News Gothic Ronaldson 0. S. Average 157.5 a 174.2 a 183.3 a 190 a 169.2 177.0 223.3 b 195.8 b 220 b 245 b 219.2 217.8 181.7 c 182.5 c 197.5 c 221.7 c 190.8 193.8 276.7 d 253.3 6 246.7 d 256.7 d 245 254.3 175 e 170 ^ 191.7 e 191.7 e 165 173.5 260 f 205 f 225 f 263.3 f 235 233.0 218.3 g 196.7 g 202.5 g 258.3 g 198.3 220.8 236.7 h 205 h 220 h 235.8 h. 232.5 222.7 219.2 i 214.2 i 231.7 i 249.2 i 204.2 224.1 210.8 1 236.7 210 J 280.8 3 247.5 239.4 206.7 k 215 fc 230 k 233.3 k 224.2 216.9 250 1 190.8 I 251.7 1 226.7 1 243.3 236.3 295 m 315 m 291.7 m 326.7 m 300 296.8 203.3 n 197.5 n 206.7 n 220 n 190 195.7 193.3 o 170.8 o 188.3 207.5 o 180 190.1 239.3 p 225 P 255 p 248.3 P 217.5 236.1 238.3 q 212.5 q 220.8 q 257.5 q 217.5 226.4 190 r 198.3 r 196.7 r 237.5 r 181.7 203.6 138.3 s 154.2 s 150 s 177.5 s 141.8 152.6 191.7 t 191.7 t 188.3 t 223.3 t 202.5 199.6 195 u 194.2 « 180 u 215 u 185.8 193.1 205.8 V 199.2 V 219.2 V 235.8 V 219.2 213.1 280 w 265 w 270 w 305 w 263.3 261.6 171.7 X 185 X 192.5 X 196.7 X 170 181.7 226.7 y 208.3 y 241.7 y 246.7 y 210 224.6 175 z 170 z 171.7 z 199.2 z 185 171.6 214.2 204.8 214.7 236.4 209.2 213.7 TABLE II. LOWER CASE Showing the Average Distance, Expressed in Cm., at Which Each Letter of Each Face was Read (Twelve Readings, Six Ob,server8). The First Column in Each Division of the Table Shows the Actual Size and Form op the Letter Which was Presented for Identification; and the Number Indicates the Average Distance, From the Ete, at Which the Letter was Correctly Identified American Typewriter Bulfineh Caslon 0. S. Century 0. S. Century Expanded Cheltenham 0. S. Cheltenham Wide Clearface Gushing 0. S. Cashing No. 2 Cushing Monotone Delia Robbia De Vmne No. 2 Jensen 0. S. News Gothic Ronaldson 0. S. Average a b d e 184.2 208.3 168.3 230 162.5 a b c d' e 185 233.3 207.5 258.3 181.7 a b c d e 170.8 190 185 253.3 165.8 a b c d e 197.5 245 230 199.1 a b c d e 196.7 233.3 273.3 187.5 a b c d e 155 221.7 180.8 261.7 169.2 a b c d e 190.8 245 200.8 276.7 185.8 a b c d e 182.1 223 235.9 283.4 179.8 a b c d e 173.3 206.7 179.2 260 164.2 a b c d e 155.8 171.7 174.2 205. S 140 a b c d e 155.8 203.3 171.7 233.3 146.7 a b c d e 167.5 223.3 181.7 276.7 175 a b c d e 174.2 195.8 182.5 253.3 170 a b c d e 183.3 220 197.5 246.7 191.7 a b c d e 190 245 221.7 256.7 191.7 a b c d e 169.2 219.2 190.8 245 165 177.0 217.8 193.8 254.3 173.5 f g h i 3 241.7 203.3 207.5 206.7 256.7 f i i 261.7 251.7 241.7 243.3 228.3 f S h i i 225 182.5 210.8 209.2 235 f r 1 i J 256.7 245 235 245 240 f g h 1 J 256.7 241.7 238.3 215.8 283.3 f I i i 235 195.8 211.7 235 234.2 i S h j 241.7 216.7 238.3 251.7 248.3 f i j 241.7 250 247 213 264.4 f e h i j 232.5 231.7 211.7 231.7 227.5 f g h i J 184.2 225.8 190.8 200 218.3 f g h i j 178.3 215 200 215 208.3 f i j 260 218.3 236.7 219.2 210.8 f g h i j 205 196.7 205 214.2 236.7 f i i 225 202.5 220 231.7 210 f g h j 263.3 258.3 235.8 249.2 280.8 f g h i j 235 198.3 232.5 204.2 247.5 233.0 220.8 222.7 224.1 239.4 k 1 m n 188.3 230 195 195.8 188.3 k 1 m n o 236.7 256.7 336.7 213.3 220 k 1 m n 197.5 219.2 291.7 190 180.8 k 1 m n 240 245 310 213.3 211.7 k 1 m n 235 252.5 286.7 199.2 213.3 k 1 m n o 211.7 240 295 181.7 181.7 k 1 m n o 224.2 254.2 315 208.3 189.2 k 1 m n 223.2 280.2 325.1 213.7 199.5 k 1 m n 221.7 232.5 315 172.5 172.5 k 1 m n 183.3 208.3 273.3 162.5 170 k 1 m n 200 199.2 276.7 163.3 175 k 1 m n o 206.7 250 295 203.3 193.3 k 1 m n o 215 190.8 315 197.5 170.8 fc 1 m n 230 251.7 291.7 206.7 188.3 k 1 m n 233.3 226.7 326.7 220 207.5 k 1 m n 224.2 243.3 300 190 180 216.9 236.3 296.8 195.7 190.1 P q r s t 238.3 233.3 210.8 149.2 198.3 P q r s t 271.7 240 216.7 167.5 230.8 P q r s t 220.8 221.7 203.3 145 180.8 P q r s t 240 233.3 216.7 167.5 206.7 P q r s t 240 233.3 225 165 221.7 P q r s t 223.3 221.7 187. 5 138.3 183.3 P q r s t 247.5 219.2 214.2 160 190 P q r s t 259.7 234.4 215.4 150.4 228 P q r s t 245 235 205 145.8 208.3 P q r s t 205 196.7 175.8 145.8 169.2 P q r s t 201.7 206.7 183.3 145.5 179.2 P q r s t 239.3 238.3 190 138.3 191.7 P q r s t 225 212.5 198.3 154.2 191.7 P q t s t 255 220.8 196.7 150 188.3 P q r s t 248.3 257.5 237.5 177.5 223.3 P q r s t 217.5 217.5 181.7 141.8 202.5 236.1 226.4 203.6 152.6 199.6 u V w y z 198.3 195 183.3 180.8 230 159 u V w X y z 225 240 271.7 211.7 245 196.7 u V vv X y z 185 180.8 243.3 171.7 221.7 162.5 u V w X y z 212.5 230 266.7 200 230.8 182.5 u V w X y z 198.3 217.5 268.3 195 218.3 171.7 u V w X y z 177.5 205 263.3 169.2 222.5 164.2 u V X y z 197.5 223.3 293.3 184.2 236.7 179.2 u V w X y z 207.4 225.4 296 178.1 237.5 171.4 u V w X y z 175.8 225.8 281.7 187.5 227.5 157.5 u V w X y z 176.7 183.3 201 157.5 194.2 156.7 u V w X y z 165 205 233.3 155 195.8 143.3 u V w X y z 195 205.8 280 171.7 226.7 175 u V w X y z 194.2 199.2 265 185 208.3 170 o V w y 2 180 219.2 270 192.5 241.7 171.7 w X y z 215 235.8 305 196.7 246.7 199.2 u V w X y z 185.8 219.2 263.3 170 210 185 193.1 213.1 261.6 181.7 224.6 171.6 Average201.7 233.6 201.7 228.0 226.7 206.4 224.3 229.5 212.6 185.6 190.6 214.2 204.8 214.7 236.4 209.2 213.7 LEGIBILITY OF DIFFERENT FACES OF PRINTING TYPES I3 TABLE III. ISOLATED LETTERS A Comparison of the Average Legibility of Various Faces. Order of Legibility UPPER CASE Lower Case The Sixteen Roman Faces JENSON 281.7 News Gothic 236.4 BULFINCH 273.8 Bulfinch 233.6 CHELT. W. 268.5 Clearface 229.5 CENTURY 0. S. 270.4 Century 0. S. 228.0 CLEARFACE 269.3 Century Exp. 226.7 CHELT. 0. S. 268.5 Chelt. W. 224.3 DELLA ROBBIA 266.8 Jensen 214.7 NEWS GOTHiC 264.6 Delia Robbia 214.2 CENTURY EXP. 264.8 Gushing 0. S. 212.6 CASLON 0. S. 250.7 Ronaldson 209.2 GUSHING 0. S. 247.6 Chelt. O. S. 206.4 DE VINNE NO. 2 243.2 De Vinne No. 2 204.8 RONALDSON 241.7 American Typewr. 201.7 GUSHING MON. 228.4 Caslon O. S. 201.7 GUSHING NO. 2 224.8 Gushing Mon. 190.6 AMERICAN TYPEWR. 196.8 Gushing No. 2 185.6 Average 252.8 Average 213.7 Bold Faces CENT. O.S. BOLD 296 CHELT. O.S. BOLD 286.2 CLEARFACE BOLD 273.7 Average (Bold) 285.3 Cent. O.S. Bold Chelt. O. S. Bold Clearface Bold Average (Bold) 255.1 233.4 230.5 239.7 Italic Faces CLEAR. ITALIC 274.3 Clear. Italic 231.2 CHELT. ITALIC 259.6 Chelt. Italic 203.8 DE VINNE ITALIC 235.5 De Vinne Italic 201.9 Average (Italic) 256.5 Average (Italic) 212.3 Average (of same Average (of same faces, Roman) 260.3 faces, Roman) 213.6 Bold Italic Face CLEAR. B. ITALIC 265 . 4 Clearface B. Italic 213.2 Extra Bold Faces BOLD ANT. 307.4 FRANK. GOTHIC 284 8 Bold Antique 260 . 5 Franklin Gothic 245.2 LEGIBILITY OF DIFFERENT FACES OF PRINTING TYPES I3 TABLE III. ISOLATED LETTERS A Comparison of the Average Legibility of Various Faces. Order of Legibility UPPER CASE Lower Case Tht I Sixteen Roman Faces JENSON 281.7 News Gothic 236.4 BULFINCH 273.8 Bulfinch 233.6 CHELT. W. 268.5 Clearface 229.5 CENTURY 0. S. 270.4 Century 0. S. 228.0 CLEARFACE 269.3 Century Exp. 226.7 CHELT. 0. S. 268.5 Chelt. W. 224.3 DELLA ROBBIA 266.8 Jensen 214.7 NEWS GOTHIC 264.6 Delia Robbia 214.2 CENTURY EXP. 264.8 Gushing 0. S. 212.6 CASLON 0. S. 250.7 Ronaldson 209.2 GUSHING 0. S. 247.6 Chelt. O. S. 206.4 DE VINNE NO. 2 243.2 De Vinne No. 2 204.8 RONALDSON 241.7 American Typewr. 201.7 GUSHING MON. 228.4 Caslon O. S. 201.7 GUSHING NO. 2 224.8 Gushing Mon. 190.6 AMERICAN TYPEWR. 196.8 Gushing No. 2 185.6 Average 252.8 Average 213.7 Bold Faces CENT. O.S. BOLD 296 CHELT. O.S. BOLD 286.2 CLEARFACE BOLD 273.7 Arem^e (Bold) 285.3 Cent. O.S. Bold Chelt. O. S. Bold Clearface Bold Average (Bold) 255.1 233.4 230.5 239.7 Italic Faces CLEAR. ITALIC 274.3 Clear. Italic 231 2 CHELT. ITALIC 259.6 Chelt. Italic 203 8 DE VINNE ITALIC 235.5 De Vinne Italic 201 9 Average (Italic) 256.5 Average (Italic) 212 3 Average (of same Average (of same faces, Roman) 260.3 faces, Roman) 213 6 Bold Italic Face CLEAR. B. ITALIC 265.4 Clearface B.Italic 213.2 Extra Bold Faces BOLD ANT. 307.4 FRANK. GOTHIC 284 8 Bold Antique 260.5 Franklin Gothic 245.2 14 ROETHLEIN TABLE IV. ISOLATED LETTERS The Relative Legibility of the Letters of the Alphabet. (Average of Sixteen Faces; 192 Readings of Each Letter) UPPER CASE Lower Case W 300.2 m 296.8 M 293.8 w 261.6 L 291.1 d 254.3 J 287.5 j 239.4 I 280.4 1 236.3 A 272.4 P 236.1 T 268.9 f 233.0 C 265.1 q 226.4 V 263.5 y 224.6 Q 261.7 i 224.1 P 257.9 h 222.7 D 254.3 g 220.6 254.0 b 217.8 Y 252.1 k 216.9 U 251.3 V 213.1 F 241.6 r 203.6 H 240.5 t 199.6 X 239.8 n 195.7 G 237.6 c 193.8 N 235.5 u 193.1 Z 233.8 190.1 K 231.7 X 181.7 E 223.9 a 177.0 R 214.0 e 173.5 B 208.9 z 171.6 .s 205.7 s 152.6 Average 252.8 Average 213.8 LEGIBILITY OF DIFFERENT FACES OF PRINTING TYPES I 5 TABLE V. UPPER CASE The Effect of Various Modifications of a Given Face Cheltenham RnlH Ordinary ^ Bold Condensed Wide Italic A 291.7 A 303.3 A 190.0 A 273.3 B 210.0 B 236.7 B 204.2 Co B 220.0 C 290.0 C 295.0 C 253.3 C 283.3 D 270.0 D 284.2 D 260.0 "S D 261.7 E 259.2 E 273.3 £ 225.8 "1 E 237.5 F 241.7 F 281.7 F 239.2 F 235.0 G 275.0 G 280.0 G 232.5 G 240.0 H 256.7 H 295.0 H 213.3 H 263.3 I 268.3 I 306.7 I 273.3 I 270 J 281.7 J 295.0 J 290.0 ~ r 0.1 t^ J 276.7 K 254.2 K 277.5 K 234.2 . Widi of thii K 240.8 L 300.0 L 311.7 L 280.0 L 288.3 M 303.3 M 321.7 M 295.0 i S M 301.7 N 261.7 N 270.0 N 216.7 •li 15 TV 256.7 278.3 O 295.0 243.3 O 280.0 P 270.8 P 293.3 P 240.8 P 286.7 Q 291.7 Q 300.0 Q 258.3 ^■l Q 286.7 R 255.0 R 253.3 R 205.0 R 244.2 S;i 215.0 S 228.2 s 187.5 ? lettef Style S 206.7 T 271.2 T 306.7 T 271.7 T 258.3 U 266.7 U 298.3 U 240.0 ^1 U 263.3 V 288.3 V 305.0 V 254.2 V 240.8 W 306.7 W 323.3 w 300.0 ^^ W^ 278.3 X 263.3 X 270.0 X 255.0 X 242.5 Y 249.2 Y 273.3 Y 256.7 Y 245.0 Z 261.7 Z 263.3 Z 235.0 Z 268.3 Average 268.5 286.2 248.2 ^ 259.6 i6 ROETHLEIN TABLE VI. LOWER CASE The Effect of Various Modifications of a Given Face Cheltenham "RnlH Ordinary Bold Condensed Wide [talic a 155.0 a 195.8 a 167.5 a 190.8 a 159.2 b 221.7 b 221.7 b 209.2 b 245.0 b 195.0 c 180.8 c 198.3 c 185.8 c 200.8 c 192.5 d 261.7 d 285.0 d 275.0 d 276.7 d 263.3 e 169.2 e 185.8 e 170.8 e 185.8 e 165.0 f 235.0 f 244.2 f 214.2 f 241.7 / 261.7 g 195.8 g 222.5 g 202.5 g 216.7 g 216.7 h 211.7 h 251.7 h 238.3 h 238.3 h 241.7 i 235.0 i 226.7 • I 206.7 i 251.7 i 251.7 i 234.2 • J 280.0 • J 246.7 J 248.3 J 222.5 k 211.7 k 225.8 k 244.2 k 224.2 k 217.5 1 240.0 1 270.0 I 231.7 1 254.2 I 230.0 m 295.0 m 301.7 m 273.3 m 315.0 m 273.3 n 181.7 n 217.5 n 175.8 n 208.3 n 198.3 o 181.7 o 192.5 160.0 o 189.2 169.2 p 223.3 p 255.0 P 225.8 p 247.5 p 223.3 q 221.7 q 255.0 q 205.8 q 219.2 q 196.7 r 187.5 r 223.3 r 199.2 r 214.2 r 205.8 s 138.3 s 140.8 s 135.8 s 160.0 s 131.7 t 183.3 t 222.5 t 208.3 t 190.0 t 192.5 u 177.5 u 217.5 u 170.8 u 197.5 u 179.2 V 205.0 V 263.3 V 208.3 V 223.3 V 158.3 w 263.3 w 300.0 w 261.7 w 293.3 tt 245.0 X 169.2 X 190.8 X 185.0 X 184.2 X 174.2 y 222.5 y 245.0 y 224.2 y 236.7 j; 174.2 z 164.2 z 175.8 z 154.2 z 179.2 z 160.0 Average 206.4 233.4 205.9 224.3 203.8 LEGIBILITY OF DIFFERENT FACES OF PRINTING TYPES 17 t^ t>; C^ CO t> l> CO l>; CO CO CO .-; CO t^ »0 !>; 00 1— J T-H lO -*' CO CO CO CO rH »0 t^ 10 00 00 00 lO o 05 CO CO (M' i-I ic 05 (M 00 I> '^ »0 ^ lO (32 '^f i-H 'rtH I^ !>. t^ (N '-H 00 CO t^ i-H ut) CO lO :>p6S -g^^cb o o o -: 0)^00 o 2 bOp^ c^O i5 -2 3^ g CU el c3 oj :S 0000:2; Bulfin Clearf Chelt. Jenso] Centu: News News Delia Chelt. Cushii Clearf Centu News Delia Jenso: News . (Z) CO 00 u-on:^H-, iiijSZO clOq^c/^H D>^X>HM C^ !>■ !>. (M !>. CO CO 1> (N t^ 00 00 t^ !>; t>; CO I> CO r-I 06 CO 10 >0 i-H CO i—H T^' lO CO 00 00 T-i LO Tt^ CO CD C5 i-I CD CO CO CO Oi (M 00 t^ 'Tt^ CO CO CO Gi lO O T-H lO t^ CO 00 CO r-< CO t^ 00 T-i «J0 CO to (N Ca C^ C^ (M (M (M (M (M CO (M CO CO (M (M (M (M o b^ ^op o o Delia Clearf Clearf Centu Centu Centu Centu Clearf NeAvs Centu Chelt. Bulfin Bulfin Centu Chelt. Centu Clearf Jensoi News News g5| g^^ c! « g d-ajlJ ^X>N (N !>. 00 CO t^ t^ CO CO CO CO !>. CO CO CO Ca »0 !>; ci lO T-H 10 CO CO »o CO 00 00 00 CO CO T-H LO CO "jO CO 00 00 J> CO CO (M C5 t^ "^ CO t^ CO rH lo o '— 1 lO cri 00 O CO C^ C5 1^ CO >— 1 ^ CO i— 1 CO ] CO (M ^X><^s^ CX) CO !>. c» !> CO !>. »0 !>; 00 1> CO t^ !>. t^ 00 CO t>. CO CO 1-H 10 >0 CO '-^ C^' r-H r-H ,-H CO CO lO O »0 lO »0 1— 1 '— 1 O 00 r- 1 CO C^ GO ^ t^ t^ t^ (N CO '-< (M CO C5 C2 Ci 'f 1 :i^d |dd ^■^6 ^ tury vs Go finch tury son Clearface Chelt. 0. Jenson Delia Ro Jenson finch son laRo It. 0. tury o o 03 --^ r^ CO M >- "t; ■^ la Ro la Ro tury finch .finch !lt. 0. ^X>-N 1>. (N l>- !>. CO (M T-H !>. i> CO •>: <^ t>. CO CO !>. i-H ci CO CO CO t^ CO O »^ CO O CD T-I 00 >0 CO 00 CD O CO CO T-H O CO '— 1 >o t^ t^ 00 Oa (M !>. rH O^ t^ C5 O O to CO o t^ ^ ^ CO t^ t^ CO (M CO CO C^ (M (M . !>. t^ CO CO !>. CO to to 'i^ CO CO '^ tq !>; l>. lO CO O O to CD i-H CO lO 00 TjH i-i to CO CJ 1>I to to O GO 00 to O cm' CD T-H 00 Ol O O 00 to ^ CO CO "^ CM to 1-1 o c:5 ^+1 CO i-H CD O Ol CM 00 C5 CO t^ T-H (M (N (M 1-1 iM 3 > ^ « >> »4 CO LQ CO lO t>. CO CO t^ CO to I> to CO CO !>. 1> '^ oo J>; CO tq CM W o CO t^ CO t^ CD LO CO CO CD CO cm' to CO i>^ CO (Z5 CO CO r-J i>^ to i-J CO i> oi i-q Oi (M O t^ (X) lO "* CO '^ to CO to T-l T-H O Tt^ CO 1-H O C-1 O CM 00 O CO t^ fa T-H (N (M (M rl (>J (M WO . s ^, dd^ ,, . O O C ^ ^ op^l bb o O) ;3 J3 S -+J -(^ ;ii cC . ooom<5 ob^^o <;oooi2; ps S C c ri ;2;Gooo 1^ o <^ t>- OoO^OQ (OjD O'XS O) «w bjO-fl.--'-3 ^^ S c o Q. O" V-l M 4-3 3 > ^ X >» N 02 00 !> t^ !>. CO TtH CO CM --H CO t^ t^ tq CO tq CO t^ tq O lO i-H O >— 1 O O 00 LO '#' CO •^' to CO i-H to O CO t^ CO CM O CO to 1-5 CM Q CI '^i 1 C O O O O o oi o H-l C3 ^ fH CS ^^1 bo -+J -(^ --; +f tn -*J P5 g rd O (T) o s ^ O) D > ^ X >«N 1> !>. !>; I-^ !>. !> (M CO !>; !>. CO CO t^ j:^ to 00 t^ CO »0 O CO '— < 1— i 1— 1 1— I Ci o to GO CD CO CO O O CO t^ 00 to to CD O to CO P4 Ci ^ CO 1> C5 O to ^ '^i O) CO to O-l i-H .— 1 to to CM CO CM 1— 1 CO o o Tti oi ^ (M C^ iM ^ (N (M ^ X >.N lO Ci ^ lO CO CO t> CO I>- O-l CO CO CM to to to oo t^ I>. iq o t^ uo lO CO l>I CO 00 t^ i-I CO CO O CO O 00 i-H t^ l> I>I O to O to T— J o l>^ o o "* CO CO Oi CD to Tt^ to 00 CO 00 CO CM ^ l> to CO 1> CO CM '^ O 1— 1 m OS r-l (M (M C^ ^ (M CM (M (M (N CM CM CO CM CM CM CM CM T-l CM CM O-l CO CO CM 'th W OJGC CO _o _o &I o _o .2 .2 ^o _o _o ^:S .W do d 13 3 S S -1^ -tj -t-i 'S 'S 'S >-H ?H ^ Co ?-4 CO 5^^ b -1 ?J rH O ,H -3 ^30-3 _-. O O 0_cH oOOO o _rJ _c! O r-. O O o oOoOO ooooo M cc 't; ^J _g t^ !^ 5^ 3 c £ rfS Ps 3 M M W a ss is P P 3 Ji :3 o 3 13 O O OJ Ci 3 C! O 3 OJ O) l2;:2;ooo pqoPQ^W pq^^^W pqW^pQlz^lz; C^^ (JTS 0) H- ClD^ .rt-r-s _i^^ E c CXCTi- trt-^ 3 > 5 X >»N LEGIBILITY OF DIFFERENT FACES OF PRINTING TYPES 19 Eh [a hH ^ CJJ >H H e! hH P 1-J l-H C) PQ M S la u y^ Fi M ta H-I Ph b U) w « Q W <) 1^ p:; l p cc is CO 12: p^ IS 02 W w &H CO t^ C<1 !>. CO CO 00 CO t^ »0 00 1> !>■ CM !>. CO CD C5 CD 05 -^ rtn CO T-H 06 CO »o 10 10 06 i-H t-^ 06 i-H CO cri CO CO 10 Tt" CO rH CO t^ 00 (M CO C3i C^ CO TfH TtH Oi 10 CO 'rt^ CO Tfl i-H (M (N (M (M (M (M iM (M "^ =^ ^1> 1^ CO P5 g CO rn ^ !n tl o3 ^ o3 ^r; -^^ »C pi^ Pi^-o;^ ^ c3 9 ^ 3 -s CD q ::; CD q 3 i 13 pi 0^ ^ a; jH C^OP^P^O OOQP^P^ u^p\^6 OOOPPh P^QP^QOQ ^X>Htsj LO (M (M CO CO iC CO (N CO (M CO !>; !>; CO !>; I>; 10 1> CO l>- !>; CO »>• 1>. (>; '^^IH as CO 06 i>l CO tH oo T-H CD l-H 10 CD CO »-H CD t^ i-H CO CO i-H 06 i-H 10 10 '* T— 1 1— 1 1— 1 CD !>• C^ 00 t^ C^ (M CO CO (Ti 02 10 CO Th Oi CM "* 1-1 (M (M (M (M :a 2 :a > CC g CO M '^ •p 'Pl 03 ^3 03 CO CO CO '-' CO i-l CO CO CO 02 ,, =1 :3 =:^ :3 13^ ;3 OPhOOO 03 03 ;3 li 0^ Pi3 P2 03^ OPhOOG pi P3 03 j3 q OOOp:?OP^ <:3pqoOW Pr^OW . ^ J^^O 0. OQi coH tD>^ ><1>-N I>. !>. i>. t^ CO I> 10 00 t^ !>. »0 I>. CO (M i>. 05 CO CO >o 10 >0 CO 10 CO to i-H CD T-H CD CO r^ CO CD CD I>^ CD CO Id C5 CO T)H CO CO t^ CM t^ 06 CM CO r^ Ttl C^ LO CO r-H C3 »— 1 "O CD 1 1 CO t^ ' 1 T— ( CO (M Oi 00 CO CO CO 00 0-1 CO C^ T-H (M (M .-H C^ r-H (M C^ (M (M (M C^ (M -?- 15?- 1^ ^ ^ ^ ^ ^ ..C>->^^ ^ ^ ^ ^ ^ ^ 02 CO M 02 ^, r-H m M m CO C» ^,, ^^ CO CO M CO CO M CO CO := :3 3 s c3 P3 P3 P3^ ^ ^ ^ ^ ^ J- OJ (D ^j ^ uQQoo 2 P3 =) Pi Pi ;3 000000 <:fqun w t^ Wi-H «-? ^^Szo pL, acsf 02 H ;d>^x!>m (M_ (M_ CO CO CO CO c^ 10 (N !>. CO !>. (M_ t^ ^ 00 l> 10 CM '*' oi CO 06 CO (X) '^ C^' Tj^ 1-H 06 T-H 10 oi CD od cr> CD 10 CM "O '^ CO t^ '^ '-H 02 OO ^ CD CD CD --H (M (N CO CO (M CM CO 00 1-1 CM C^ '-H , §^ §^^ Ronaldson Gushing N Am. Type^ Gushing N Am. Type'' bC n bO bO bjO bO bC^ bC bJO be bO bfi bJO Pi bjD .s.a.s.SH q _g c g g Pi s j=; Pi Pi CO CO g CO CO g Pi fl pi.g Pi ^ ^ ^^ . ;5 j>i3ia la ;j:i "^ ^ S >■ s CC 02 M W rH CO CO CO CO CO CO CO CO , CO ;= =5 q 13 G oooo-.H^ M jSZO P^Opc; c/)EH D>^X>.N t> t^ CO 10 !>. . 10 CO Ol (M_ I> »0 CO CM 00 10 t>. iq T-H CD >0 CO (N '-H ci CD t^ 06 c^i ^' CD I> CO oi C5 10 10 cm' i-H t^ (M t^ C35 (X. Oi C3i CD C5 CO (M t^ ■* 00 CO CO CJi l-H CiO CO l-H CO C32 CM 00 C^ 1— 1 1— ( r— 1 1 — 1 -— 1 --H r-H (M (M 1 — 1 C<) 1 — 1 1 — 1 1 — 1 l-H CM l-H 1—1 CM CM CM i-H i-H CM 1— 1 ',^ u u ;h C3 U U U U lA ;^ ;^ ^^ !L^ ;^ . . • . .>>>.>.«) >>>.>.>.>. >.>>>>>> >^ >, t>,bC>,bX) >> >-. >^ >-. >i >^ a s a s 1 EhHHHH HHHHH HH.an.a HHHEhHH a a a a a a a a a a a a a a a a pi P 2 pi CO 5-1 I-H ;3 K ^ <5<:<<^}>HlSl 20 ROETHLEIN ^ H , ^ U u g M jyj (H Fh Ph t-l & 1-1 O P5 HH o i=( 1— 1 p^ ^ o w p^ Q w i, Pi CJ H « ^ (H Q Q hJ ;?; hJ n U C) O w w CO o M ffi ^ ^ <1 <; W Ph ^ ij <1 pq CJ HH O W W W H-1 H F-i cc -< H O K xn W w H CO H-) ^ o o o Z o P u CO o w o w i:^ H (M l>. l> l>. to t^ (M to 1^ I> !>. CO CM to CO CO i>- CO to 05 CO 1-5 CD to to CM o ^' t^ CO 1—1 1— 1 o o CO cJ o to o6 O to 00 1-^ 00 CM CO O 00 rfH CO CI O 1-H i-H C^l O C^l Ci 05 00 CM T-i O '^i 00 00 o -^ t^ 1-1 CO i-l (N t-H (M i-H CM CM CM CM CM CM CM CM 1—1 1—1 CM CM i-< r- 1 T-4 r-H CM C^l 1-1 CM T-I 02 aj CM m CM CM Idson 0. ng 0. S, Robbia n 0. S. Idson 0. nO. S. n 0. S. nO. S. nne No. ng 0. S, Robbia Gothic nO. S. nO. S. Idson 0. . 0. s. Robbia ng No. ! nO. S. nO. S. . 0. s. nne No. iry Exp, nO. S. Rona Cushi Delia Jenso Rona Casio Jenso Casio DeVi Cushi c3 g^^ % "^ >T^ m m ^ u- b*0j2 — .^ .bi^ £ c o D< cr* V- w -to 3 > ^ X >J tsj »0 CO (N M CO to C-1 CO CM !>. t>. 00 to CM CO CO to CM CO t^ CM i>^ CO oj to ^' to cc i>l oi 00 O oi CO i-i to O t^ O CO CO !>; ci CO o o ai CO O t^ Tt< CO CM Oi O O ^ o 1-1 oo 00 i> CM 1-H 00 Tf< 00 i> o; ■* i> 1— 1 to r-H (M i-H (M 1-1 CM T-i CM CM CM C^l CM CM i-H 1-1 C^l CM 1-1 rH T-( 1—1 1—1 CM r^ CM 1—1 . .^ . «? CM CO ,0 bJD d gcjdo • a; ^ o o Q, O fi -^^ ^ >; shin slon ntur elt. shin ^2 . rt . — ^ «2 W g CO Jenso Rona Am.' Casio Cushi o 03 +2 c> ;i^ ;ii JC o c: C3 "^ «^ Ts "m S S d CD 13 -3 O 3 QOOPhO :i 03 S -d 13 ooooo OPhOOO ^ ^ c3 Q o c OPOp:;p:?< «5 ^ o Ti o M-l fajOrCj .- ._ ^ _ S c o Ol O* W, M .faj 3 > ^ X l>.N GO 00 (M_ CO to 00 t^ to CO t^ to to to to CO 00 00 1> CO CM CO to to id '^ CO c 00 CO CM CM l>^ CM CO i-I O CD to CO Ci 0C> t^ to C5 t^ CO CO O Ci O O 1-H C5i O t^ t^ I> T-I l-H (^ Th 00 i> cn CO CO o to 1— 1 T— 1 1 — 1 (N i-H CM >-< CM CM CM 1—1 CM CM 1-1 1-H CM CM 1-1 1-; 1-1 T— 1 T— 1 CM 1— < CM 1— < .^^1 . . CM CM CM . . . CO <^. .C/2 ^ . ^. . Co . C !-> d d ;-*' • . . flCCCC Ison ne No g Mon [son 0. S. . f^ a c o2 s '^ >; ^ CD hJ3 d bJD bC >^ neN 0. S. neN ypew 0. S 0)00.. ^!2;^oo bjj >, bX)0 G bX) Cushin DeVin Cushin Cushin Am. T; DeVin Chelt. DeVin Am. T; Jenson «.s .s.s.s ^ ^ !5 S ^ CO CO CO 02 CO nald :Vin shin nald slon shin .1. T; shin elt. Vin shin c3 ;3 d d S ooooo O 0^ ^3 O 03 p:5P0P40 o<:ooPo cj .D o t::! d) <^ OOJS-H .^ j^^ S d o 0< O" n C/2 +-I P > ^ X >> N 00 !>; t^ CM t^ CM !>; CO 00 CO CO 00 t^ t^ CM 00 00 GO to GO 1> to O i-H O CO "^ i-H O Tj^" O 00 cJ CO ■*■ o to CO i-J "^ ci to O O t>^ to CD to Oi t^ CO '^l 00 C5 o o ^ 00 C5 t^ CO t^ o o 00 00 t^ t^ (X O to Cj: to 1 — 1 1— ( r— 1 C^ T— 1 1-1 1— 1 CM C-1 CM T— i 1 — 1 CM »— 1 1— 1 CM CM 1—1 CO 1-1 1—1 1—1 C-i 1—1 1—1 1—1 (M (N^ M_ . •^(M .^. <^^ ■^■'2 • CM CM .CM . . c ^ c ing No. inne No. ing Mon Mson 0, 1 Robbia ^* d .Co . co.^ fi CO . . . g . oco o ^ O bJo'O bC >" bD >. fl biD bJO fl ing No ing Mo ildson < . Robb: ing Mo . CO c o o O bjo'^' bC bJO bJD bC •S g.SH.S H.S.S.S.S c c f^ c c c rd ^ rd . ^ W CO M rH 03 13 ci ;3 G ;3 CO "^ 03 K —1 S a; g ^ a; CO -f: e '— < ^ 03 03 CO CO CO 02 ^ O j3 O O) 3 US q ;r > ^ X >, N i>. CO 00 CO to 00 CO CO to tq 00 !>. 00 CO CM CO CO CM CO to i-H a5 to o 00 CM* O O CO CO t^ to c^i o O C5 t^ CO CO CO CO X to o -^ 1 — 1 1 — 1 1 — 1 O'l 1— 1 1-H r-( 1-1 CM C^l 1 — 1 1 — 1 1 — 1 r— 1 1— 1 CM »— 1 1 — 1 1 — 1 1— 1 T-H 1-H T— t 1— 1 1—1 1-H (M . , bC bJO O O O ./ o c o ^ o o o O bjD >^ bC biD S .^^S Slz; |SJz;§ bc'^ bJO bC bC bC bC bcO bJD bC bC >^ bC bC b£i ^.SH.S.S .S rt.S.S s _g S H .S .S C C C . C c c ^ c c g ;i^xj .AA rC^-d^^ IS ^ . ^ !5 '^'^'rs:i;i={'^ s;h .is:s;j3

^ X >» N LEGIBILITY OF DIFFERENT FACES OF PRINTING TYPES 21 TABLE XI. THE AVERAGE LEGIBILITY OF GROUPED LETTERS The Numbers in this Table Indicate the Average Distance AT Which Each Letter of Each Face Was Read. Nine Face's Were Selected for this Experiment; and only Lower Case Letters Were Employed. The Averages are Compiled from 'Internal' Letters Only, i. e., the Initial Letters and the Final Letters of the Groups are not Included in these Averages a; o d o -^ o e (i, !Xl (^^ c ai o c3 X 6 a o 6 s c c^ -2 1-1 CI a bJO fcJD O O m »— « "m c C o m 2 > o ^ a QJ o; I-* 3 r-"