Of Interest to Publishers 
 
 A Rational System for 
 Measuring Composition 
 
 BALTIMORE 
 
 WILLIAMS & WILKINS COMPANY 
 1913 
 
Of Interest to Editors 
 
 This paper was written for the printer and the arguments advanced 
 are from the printer’s standpoint. It has always been our policy to 
 enlighten our customers on such technical points which are of interest 
 to those who publish, and a knowledge of which often proves of intrinsic 
 value to those who are buying printing. 
 
 We believe that an understanding by the customer of the difficul¬ 
 ties to be surmounted in the production of a high standard of book work 
 will lead to a greater appreciation of the finished product, and a greater 
 effort on his part to effect real cooperation with the printer. 
 
 This same subject will be presented within a short period, advancing 
 the arguments from the publisher’s standpoint. Keep this pamphlet 
 until the second one reaches you, then re-read it after reading the sec¬ 
 ond one, and we believe that you will understand that the real interests 
 of both parties to a contract, are practically identical; that mutual confi¬ 
 dence will lead to a degree of cooperation which will mean increased 
 efficiency, bring a higher value in workmanship and service to the 
 customer, and more uniform and probably somewhat better profits to 
 the printer. 
 
 We have demonstrated to others that this is so and we are prepared 
 to substantiate our claims to you. 
 
 THE WAVERLY PRESS, 
 
 Williams & Wilkins Company, Proprietor, 
 Baltimore, U. S. A. 
 
A RATIONAL SYSTEM FOR MEASURING 
 COMPOSITION 
 
 EDWARD B. PASSANO 
 
 (President of Williams & Wilkins Company , Baltimore) 
 (Reprinted from the Inland Printer for July) 
 
 <4 
 
 i 
 
 P 
 
 P 
 
 Speaking broadly, cost records are kept for two purposes: 
 First, to be able to tell before a job is taken what price to charge 
 to make a profit on it; second, to be able to tell after the job is 
 delivered whether a profit was made—whether the estimate was 
 right. An estimate that is too low means the loss of money; an 
 estimate that is too high means the loss of the job and the money that 
 might have been made out of the job. 
 
 Now the prime factor in estimating on anything, from putting 
 carpets on floors to putting printing ink on paper, is determining 
 the amount of work that must be done. A man who could not 
 measure accurately would not long hold a job in a carpet store, 
 where mighty little carpet is worth $5 a square yard. 
 cr > <^ > Printers have been unable to estimate accurately the cost of 
 composition, for there has been no accurate system of measuring 
 composition, and yet there is mighty little composition that is 
 not worth more than $50 a square yard. 
 
 The system, or rather complete lack of system, of using the em 
 of the point size of the face to measure composition is ridiculous 
 :ja&id comment would be superfluous but for the important fact that 
 this method is universally used. 
 
 When composition is measured with the em of the point size 
 of the face, no consideration whatever is given to whether the face 
 used is fat or lean. Type faces of the same point size vary more 
 in width than carpet does. How long would a carpet estimator 
 last who assumed that all carpet was the same width, and in figur- 
 ^ ing the number of yards to cover a floor, paid no attention to the 
 width of carpet for the job on which he was estimating? 
 
 l 
 
 4 
 
2 
 
 EDWARD B. PASSANO 
 
 A dollar represents the same amount of value whether you pay 
 the dollar in quarters, dimes, nickels or pennies. 
 
 What does a thousand ems represent? Does a thousand ems 
 of twelve point represent the same value (amount of work) as a 
 thousand ems of six point? Why, there is often a difference of 
 more than 30 per cent in the value of two different thousand ems of 
 the same point-size type. 
 
 Here are three eight point faces: 
 
 abcdefghijklmnopqrstuvwxyz {case^aijhabe?} U - 23 8 Pt - Ems. 
 
 abcdefghijklmnopqrstuvwxyz {casl^aiphablt } 12 - 83 “ “ 
 
 abcdefghijMmnopqrstuvwxyz{^\^^} 14 - 42 “ 
 
 The fattest of these faces will set almost 30 per cent faster than 
 the leanest, and the operator of a composing machine when setting 
 the fattest of these faces will produce 30 per cent more ems in the 
 same time than when setting the leanest; and this on setting iden¬ 
 tically the same matter—striking exactly the same keys. And yet, 
 in obtaining output records upon which to base estimates, how 
 many printers give any consideration whatever to whether the 
 faces used be fat or lean? 
 
 It is obvious that in selling printing, selecting the most profitable 
 face for the job is of the utmost importance, and this is becoming more 
 important as editions grow larger. In much work the saving in 
 paper and press work effected by using a suitable condensed face 
 and spacing closely adds a very handsome profit, while, when 
 composition is sold by the page, the advantage of using a fat face 
 and wide spacing should not be overlooked. 
 
 Whether, therefore, you charge composition to the completed 
 job at a rate per hour, of a rate per thousand ems, you must know 
 your average output per hour, based on a system of measurement 
 that really means something, if your cost records are to be of any 
 real value. 
 
 The prime object of a cost system is to enable you to sell at a profit — 
 to give you the data to estimate accurately on new work. It is absurd 
 to spend money on a cost system to get accurate hour costs, and 
 then, in estimating, throw away all this advantage by figuring 
 
A RATIONAL SYSTEM FOR MEASURING COMPOSITION 
 
 3 
 
 on the totally false assumption that all kinds of matter, set in fat 
 or lean faces, can be produced at the same average number of ems 
 per hour. What can be more foolish than giving the buyer of 
 printing, who says, “I do not care for quality and will not pay for 
 it, ” more words to the page than your competitor does, unless by 
 more closely setting the job you can make fewer pages and save 
 paper and press work. 
 
 It is a fundamental principle of arithmetic that before any quan¬ 
 tities are added to obtain an average they must be reduced first 
 to the same denomination; you cannot add inches to feet and call 
 the result either feet or inches; first you must express the quanti¬ 
 ties to be averaged in feet or in inches. 
 
 In the same way composition output records must be reduced 
 to a common basis before they are averaged. Proper allowance 
 must be made for whether the faces be fat or lean. How shall this 
 allowance be made? 
 
 It is clear to all who have studied this question that the generally 
 accepted standard for comparing faces, the length of lower-case 
 alphabet, is too inaccurate to be of any real help in solving this 
 problem. Indeed, in the old days of piece work on hand compo¬ 
 sition faces that looked fat by this standard but set skinny, be¬ 
 cause the most frequently used letters had been made condensed, 
 were a constant source of irritation. If scientific management 
 means scientific management, it is clear that any system of meas¬ 
 urement to be acceptable today must take cognizance of something 
 more definite than the alphabet length. 
 
 The monotype keyboard furnishes the ideal measuring instru¬ 
 ment upon which to base an accurate system of measurement 
 because this keyboard is both a measuring and adding machine. 
 As each letter and space is struck its width is tneasured and re¬ 
 corded so that, after a line is set, the keyboard indicates exactly 
 the number of ems in the line. To those not familiar with the 
 monotype system the following will be of interest: 
 
 In using the keyboard the operator sets the measure not to the 
 measure in pica ems as a compositor sets his stick, but to the 
 equivalent, in picas, of this measure in ems of the “set” of 
 the face to be composed. The conversion of pica ems to ems 
 
4 
 
 EDWARD B. PASSANO 
 
 of any set is quickly made by using a table calculated for all 
 sets. 
 
 In the monotype system “set” is the measure of a face “line- 
 wise” and tells exactly how condensed or extended the face is. 
 Referring to the three eight point alphabets, on page 2, the first, 
 or most condensed, of these belongs to a seven-set monotype face; 
 seven-set because the widest character used with this face is seven 
 points wide. The second alphabet is eight-set, while the third, 
 most extended, is nine-set. Thus, in composing matter seventeen 
 picas wide in the most condensed of these faces, the monotype 
 operator sets his keyboard measure to twenty-nine and one-sixth 
 ems because his table for converting picas into ems of any set 
 tells him that twenty-nine and one-sixth ems (widest characters) 
 of this face is exactly seventeen picas wide. 
 
 The possibilities of using the “set” of a face as a basis for ac¬ 
 curately measuring the amount of effort required to compose a 
 required number of square inches of this face will be clear from 
 the following specimens of monotype composition. Bear in mind 
 that to use this same system of measurement for hand-set foundry 
 type, or any other process of composition, we have only to first 
 determine the “set” of the face we wish to measure accurately. 
 
 In the old system of measurement (“traditional” measurement) 
 the amount of matter set is expressed in ems of the point size, the 
 result of multiplying the measure in ems of the point size of the 
 face by the number of lines set; the rational system of measure¬ 
 ment gives the product in “Set Ems,” that is, the measure in ems 
 of the set of the face multiplied by the number of lines. 
 
 These specimens also show clearly the folly of measuring output 
 with the em of the point size of the face as a “yard-stick” (“tradi¬ 
 tional” measurement), regardless of whether the face be fat or 
 lean; because, with the exception of the quads required to finish 
 the last line of the paragraph, the operator made the same effort, 
 struck the same keys, in setting each specimen; the width in all 
 cases is eighteen and one-half picas. 1 
 
 1 For the purpose of accuracy in these calculations, the fractions of the ems have 
 been considered, but the usual habit should be followed in compiling statistics of 
 not considering less than an em, dropping the fraction when less than one-half, 
 and counting the next whole em when one-half or over. 
 
A RATIONAL SYSTEM FOR MEASURING COMPOSITION 
 
 5 
 
 The first specimen, figure 1, is set in a seven-set face, the key¬ 
 board measure is thirty and one-half ems six units (J em); multi- 
 
 The best kind of originality is that which comes after a sound appren¬ 
 ticeship; that which shall prove itself to be the blending of a firm 
 conception of useful precedent and the progressive tendencies of an able 
 mind. For, let a man be as able and original as he may, he cannot 
 afford to discard knowledge of what has gone before or what is now 
 going on in his own trade and profession. If the printers of today 
 do not wish to be esteemed arrogant, when they term this calling of 
 theirs an art, they must be willing, and show that they are willing, 
 to subject it to such laws as have made its sister arts so free. 
 
 8 pt. 1A (7 set) =9 lines 
 
 Measure 18 picas (27 ems of 8 pt.); Traditional measure¬ 
 ment =243 ems 
 
 Measure 18 picas (30K ems six units of 7 set); Set Ems 
 measurement =277K ems 
 
 Figure 1. 
 
 plying this by the number of lines (9) gives 277J real ems , of this 
 face, or “Set Bms,” the name for the output measured in ems of 
 the set of the face in which the matter is composed. But according 
 to tradition this amount of matter should be called 243 ems, be¬ 
 cause eighteen picas equals twenty-seven ems of eight point and 
 the matter makes nine lines (27 X9 = 243). 
 
 The best kind of originality is that which comes after a sound 
 apprenticeship; that which shall prove to be the blending of 
 a firm conception of all useful precedent and the progressive 
 tendencies of an able mind. For, let a man be as able and 
 original as he may, he cannot afford to discard knowledge of 
 what has gone before or what is now going on in his own trade 
 and profession. If the printers of today do not wish to be 
 esteemed arrogant when they term this calling of theirs an 
 art, they must be willing, and show that they are willing, to 
 subject it to such laws as have made its sister arts so free. 
 
 8 pt. 14A (8 set) = 10 lines 
 Same matter as Fig. 1 
 
 Measure 18 picas (27 ems of 8 pt.); Traditional measure¬ 
 ment =270 ems 
 
 Measure 18 picas (27 ems of 8 set); Set Ems measure¬ 
 ment = 270 ems 
 
 Figure 2. 
 
 The next specimen, figure 2, is set in an eight-set face and meas¬ 
 ures 270 “Set Ems,” also 270 point-size ems; the same number 
 
6 
 
 EDWARD B. PASSANO 
 
 of ems by both systems of measurement because an eight-set em 
 is eight points wide. 
 
 The best kind of originality is that which comes after a 
 sound apprenticeship; that which shall prove to be the 
 blending of a firm conception of all useful precedent and 
 the progressive tendencies of an able mind. For, let a 
 man be as able and original as he may, he cannot afford 
 to discard knowledge of what has gone before or what is 
 now going on in his own trade and profession. If the 
 printers of today do not wish to be esteemed arrogant 
 when they term this calling of theirs an art, they must 
 be willing, and show that they are willing, to subject 
 it to such laws as have made its sister arts so free. 
 
 8 pt. 5A (9 set) =11 lines 
 Same matter as Figs. 1 and 2 
 Measure 18 picas (27 ems of 8 pt.); Traditional measure¬ 
 ment =297 ems 
 
 Measure 18 picas (24 ems of 9 set); Set Ems measure¬ 
 ment =264 ems 
 
 Figure 3. * 
 
 The third specimen, figure 3, is a nine-set face and, as shown 
 beneath the specimen, this amount of matter is 264 “Set Ems,” 
 or 297 “traditional” ems. 
 
 Thus , if we use the “Set Ems ” system of measurement the same 
 words , the same number of keystrokes on a composing machine , 
 make practically the same number of ems . But if we use the “ tradi¬ 
 tionalmeasuring system , based on the point size of the face , the 
 variation in measurement for the same amount of effort is as follows: 
 
 8 pt. 1A (7 set) = by Set Ems, 277§; by “tradition/’ 243 ems 
 8 pt. 14A (8 set) = by Set Ems, 270; by “tradition,” 270 ems 
 8 pt. 5A (9 set) = by Set Ems, 264; by “tradition,” 297 ems 
 
 Thus, by “Set Ems” measurement the maximum difference in 
 measurements is thirteen and one-half ems, by “traditional” 
 measurement the difference is fifty-four ems. 
 
 Before the “Set Ems” system of measurement can be used it 
 is, of course, necessary to know the “set” of the face to be meas¬ 
 ured, in order to convert any measure in picas into ems of the set 
 of the face by using the monotype table for changing pica ems to 
 ems of any set. 
 
A RATIONAL SYSTEM FOR MEASURING COMPOSITION 
 
 7 
 
 The monotype specimen book gives the “set” of all monotype 
 faces and we may determine the “set” of a type or slug face by 
 matching it with a monotype face, the “set” of which is known. 
 It is clear that if two faces set line for line on the same matter, in 
 the same measure, they must be of the same “set” and that the 
 same operator effort, the same number of key-strokes, will be 
 required to compose the same number of “Set Ems” in either of 
 these faces, whether the operator uses a monotype or a slug ma¬ 
 chine. Thus, the “set” of a type or slug face may be determined 
 by composing in this face the same matter as is used in the mono¬ 
 type specimen book—“The best kind of originality is that which 
 comes after a sound apprenticeship, etc.”—and then finding from 
 the monotype specimen book the “set” of the monotype face 
 that runs line for line with this test specimen. 2 
 
 The monotype keyboard may be used as a measuring machine 
 to determine the “set” of any foundry or slug face; that is, to find 
 from a printed specimen the “set” of the monotype face that will 
 run line for line with this face, on the same matter in the same 
 measure. The following explanation of this method for finding 
 the “set” of a face is of interest, but it should be noted that once 
 a face is measured and its set determined this work need never 
 be repeated for that face. 
 
 To find the set of any face from a printed specimen, set ten 
 full lines (no lines containing quads) on the monotype keyboard 
 to determine the number of ems and fractions of an em (units) in 
 each line. Of course fixed size spaces, not justifying spaces, of 
 
 2 In determining the set of a face by comparison always refer to monotype 
 Roman faces that are on Arrangement C. It is a simple matter on the monotype to 
 combine in the same matrix case, and therefore to compose together, a condensed 
 Roman and an extended Boldface because the Boldface matrices are placed in the 
 matrix case in positions that prbduce much wider characters than the correspond¬ 
 ing Roman letters. To facilitate making these combinations the Montoype Speci¬ 
 men book always gives, for faces not on Arrangement C, the set of the Roman face 
 with which this Boldface will combine; that is, the set of the justifying scale to 
 use in composing the Boldface shown—for example eight-point 11J, Arrangement 
 C2, Boldface is listed as an eight and one-half set face. For the purposes of meas¬ 
 urement this Boldface is in reality a nine and three-quarter set face because it 
 will compose line for line with a nine and three-quarter set Roman face on Arrange¬ 
 ment C. 
 
8 
 
 EDWARD B. PASSANO 
 
 the same width as the spaces in the printed specimen must be 
 used. From the number of ems and units in each of these ten 
 test lines obtain the average number of ems and units per line. 
 Measure the width of the specimen in picas with a type rule. In 
 monotype system, a face that has its widest character one pica 
 wide is twelve set; therefore, by simple proportion, the old rule of 
 three of arithmetic, we can find the set of the face tested on the 
 keyboard because we have (a) the average number of ems and units 
 per line, (6) the measure in picas, (c) twelve, the set of a face that 
 would give as many ems to the line as the measure in picas. We 
 can express this relation as follows: 
 
 Set of face tested on keyboard = 
 
 Measure in picas X 12 
 Number of ems to the line 
 
 Suppose that the average number of ems to the line of the face 
 tested was twenty-three and that the matter tested was seventeen 
 and one-half picas wide. Then the set of this face would be 9.13, 
 or nine and one-quarter set; the table for changing pica ems to 
 ems of any set gives the difference in faces by quarter sets. 3 
 
 1^X12 210 _ 
 
 23 23 
 
 In thus determining the set of faces it is essential to take into 
 account the necessary wide spacing of slug machine faces. In 
 monotype composition, as in hand work, the spaces between words 
 are proportional to the size of the face. In slug faces the same 
 size spaces are used with all faces and consequently the spaces in 
 six-point slug composition are twice as wide as in hand composi¬ 
 tion. Therefore in using the keyboard to determine the set of 
 a hand-set face use six-unit spaces between the words of the test 
 lines set on the keyboard; for a slug face use twelve-unit spaces. 
 
 To be fair to yourself , to be just to your operators , if you keep 
 your records in “Set Ems” never talk about your ems per hour or 
 compare your records with those kept in the “traditional” way until 
 
 3 For those who have not access to a keyboard the Monotype Company will 
 determine the set of any face, twelve point and smaller, without charge, upon 
 receipt of a page of composition in this face. 
 
A RATIONAL SYSTEM FOR MEASURING COMPOSITION 
 
 9 
 
 you have first converted your “Set Ems” into the “ traditional ” ems 
 of the point size. To do this increase your output statements by not 
 less than 20 per cent. This is essential in comparing foundry type 
 and monotype records with slug machine output because few, if any, 
 slug machine faces are so condensed that their “set” is equal to 
 their point size and also because of the necessary wide spacing of 
 these faces. 
 
 By using “Set Ems” in estimating you automatically make 
 proper allowance for fat and lean faces. A job is to be set in eight 
 point, twenty-one picas wide, sixty lines to the page. If a nine- 
 set face be used each page will make 1680 “Set Ems,” for the table 
 for changing pica ems to ems of any set shows that twenty-one 
 picas equals twenty-eight ems of nine-set (28x60 = 1680). If 
 your records show that your composing machines average 3600 
 “Set Ems” per hour (60 “Set Ems” per minute) the time required 
 for this page will be twenty-eight minutes (1680-^60 = 28). 
 If, instead of a nine-set you use an eight-set face there will be 
 1890 “Set Ems” per page (21 picas = 31J ems of eight-set: 
 31^X60 = 1890) and thirty-one and one-half minutes will be 
 required to set a page (1890^60 = 31J). Thus, your price on 
 the eight-set face should be 12.5 per cent higher than on the nine- 
 set if your composing machines maintain the same rate of profit. 
 
 But by far the greatest advantage of keeping output records 
 in “Set Ems” is that it keeps constantly before you the profit— 
 or loss—in type faces. It almost forces you to furnish the face 
 that will give you the greatest profit on the job you are estimating 
 on. If the edition be large you make an additional profit by sav¬ 
 ing on paper and press work, selecting a suitable condensed face 
 and setting it close spaced. If the composition is sold by the page 
 you use as fat a face as possible and make sure that it is wide 
 spaced. In short, if you are in competition you furnish no more 
 words to the square inch than your competitor does unless you 
 get paid an extra price for furnishing an article that costs more to 
 produce. 
 
vis wmoo' 
 
/ 
 
 30 
 
 12 072472902