pRflNKLiN Institute Librart 
 
 PHIL/lbELFHI/l 
 
 Class -^7 I Book.. X. S3. .5" Accession. 4.6 
 
 2J 
 
 Article V. — The Library shall be divided into two classes ; the first 
 comprising such works as, from their rarity or value, should not be lent 
 out, all (inbound periodicals, and such text books as ought to be found in a 
 library of reference, except when required by Committees of the Institute, 
 or by Members or holders of second class stock, who have obtained the 
 sanction of the Committee. The second class shall include those books 
 intended for circulation. 
 
 Article VI. — The Secretary shall have authority to loan to Members and 
 to holders of second class stock, any work belonging to the second class, 
 subject to the following regulations : 
 
 Section 1. — No individual shall be permitted to have more than two books 
 out at one time, without a written permission, signed by at least two mem- 
 bers of the Library Committee ; nor shall a book be kept Out more than 
 TWO WEEKS ; but if no one has applied for it, the former borrower may 
 renew the loan. Should any person have applied for it, the latter shall 
 have the preference. 
 
 Section 2. — A fink of ten cents per week shall be exacted for the 
 detention of a book beyond the limited time ; and if a book be not returned 
 within three months, it shall be deemed lost, and the borrower shall, in 
 addition to his fines, forfeit its value. 
 
 Section 3. — Should any book be returned injured, the borrower shall pay 
 for the injury, or replace the book, as the Library Committee may direct; 
 and if one or more books, belonging to a set or sets, be lost, the borrower 
 shall replace them or make full restitution. 
 
 Article VII. — Any person removing from the Hall, without permission 
 from the proper authorities, any book, newspaper, or other property in 
 charge of the Library Committee, shall be reported to the Committee, who 
 may inflict any fine not exceeding twenty-five dollars. 
 
 Article VIII. — No Member or holder of second class stock, whose 
 annual contribution for the current year shall be unpaid, or who is in 
 arrears for fines, shall be entitled to the privileges of the Library or Read- 
 ing Room. 
 
 Article IX. — If any Member or holder of second class stock, shall refuse 
 or neglect to comply with the foregoing rules, it shall be the duty of the 
 Secretary to report him to the Committee on the Library. 
 
 Article X. — Any Member or holder of second class stock, detected in 
 mutilating the newspapers, pamphlets or books belonging to the Institute, 
 shall be deprived of his right of membership, and the name of the offender 
 shall be made public. 
 
I 
 
A TREATISE 
 
 ON 
 
 EOLL-TUM/] 
 
 FOR THE 
 
 MANUFACTURE OfH^&SI^. 
 
 B-x., PIvTOR . TUNNER, 
 
 MEMBER 0t'**^^-^T3ji*i^*Vm4rj5i: oj^*;\^i*^^^ '^r^,: 
 
 TRANS LAI? E i5'* A ^^'!6'**Ab*l*t*'T E O 
 
 BY 
 
 JOHN B. PEARSE, 
 
 METALLURGIST, ENGINEER, AND MANAGER AT THE WORKS OF THE 
 PENNSYLVANIA STEEL COMPANY. 
 
 ILLUSTRATED BY 34 WOOD-OUT EN&EAVIN&S, TOGETHER WITH A FOLIO 
 ATLAS OF LITHOGRAPHED PLATES, 
 
 <S NEW YORK : 
 D. VAN NOSTRAND, Piiblishei-, 
 
 23 MURRAY ST. AND 27 WARREN ST. 
 
 1 8 6 9. 
 
AUTHOR'S PREFACE. 
 
 The use of rolls for tlie manufacture of iron is 
 unquestionably an English invention. Tlie first grooved 
 rolls were constructed in tlie year 1783 by Henry Cort, 
 wlio thereby laid the foundations of a magnificent indus- 
 try, but suffered grievous ill-usage, and died a miserable 
 death. 
 
 The subject of roll-turning has been so scantily 
 treated in all the books which have described the 
 manufacture of iron, that very httle that is useful 
 can be. gleaned from their pages. None of the late 
 books on the subject contain more information than was 
 given by Karsten in 1841, though since that time extra- 
 ordinary progress has been made in the art. The chief 
 ground of this neglect of this vital branch of the in- 
 dustry is, in my opinion, to be found in the fact that 
 those who have lately written on Iron Metallurgy 
 have not, as a rule, been practical metallurgists, but 
 only metallurgical chemists, and, therefore, they have 
 neglected as trivial such things as passes, or, perhaps, 
 have even held it beneath their dignity to write about 
 them ; they have, however, accompKshed a great deal in 
 their own branch. 
 
iv ■ author's preface. 
 
 O^duard Matirer published in 1865 a work, or ratlier 
 ,• attlas, containing tlie finished sections of many different 
 ^' kirn^ds of viron ; but there was nothing in his work which 
 treated of roll-turning except a few ideal drawings, in 
 which the finishing passes of many sections were drawn 
 on the same pair of rolls. Some years ago, Mr. Bieder- 
 man, a true metallurgical engineer, proposed to publish a 
 work on roll-turning, illustrated with correct drawings, 
 which, together with the corresponding manuscript, I 
 have seen, but the publication of which was given up for 
 reasons unknown to me. 
 
 These facts show that there must be some great diffi- 
 culty in publishing such a work, which is, indeed, actually 
 the case. This difficulty is caused by two cu'cumstances : 
 by the fact that the art rests not on theory, but on wide 
 experience — the men of experience being seldom able, and 
 rarely willing, to publish their knowledge for the benefit of 
 others; and by the fact that if such a work shall really 
 serve a useful purpose, so many drawings are necessary 
 that the cost of the work becomes excessive, and the labor 
 of the author unremunerative. The first circumstance, 
 rather than the latter, has therefore prevented me from 
 writing any complete treatise on roll-turning, though I have 
 often treated of special points, and in 1838 furnished the 
 text and drawings for a small work on rail-making, which 
 was published by Industry and Trade Society of Inner 
 Austria. 
 
 For some 25 years I have felt the necessity of such 
 a work as the present, and felt it the more deeply when I 
 observed, as I often had the opportunity of doing, that 
 many men, and especially those from abroad, regarded 
 themselves as indispensable to this or that mill merely 
 
AUTHOES PREFACE. /' 
 
 because they possessed the drawings of ?l few sets Cd 
 rolls which had been used and liked. - , . ^''r,, 
 
 I have, therefore, at last determined to pul^ish, tl^ 
 treatise, which, however, makes no pretence to ' i^falfes''^ 
 bUity or entire completeness. 
 
 P. TUNNEB. 
 
TRANSLATOR'S PREFACE. 
 
 In translating the present work I have followed the 
 original text almost literally, and have placed my own 
 additions in their proper places, and as far as possible at 
 the end of the articles to which they belong. The author 
 has written so thoroughly on every detail of roll-turning 
 that there remained little to say except on a few points, 
 the most important of which are the prevention of fins, 
 and the relative advantages of large and small rolls. I 
 have endeavored, as respects the former point, to so 
 refer in various ways to the different methods of pre- 
 venting fins as to show that, in my opinion, the most 
 important principle of rolling is to prevent them, while 
 giving all passes the requisite draw. For instance, it 
 requires the nicest skill to proportion passes for red short 
 iron, in such a manner that the bars shall be rolled so 
 quickly, and with such draw, that they may be smooth 
 at the finish, without at the same time giving so much 
 draw as to force the metal into the interstices of the 
 pass. I have also advocated the use of large rolls, be- 
 cause they are stronger and less injurious to the bars 
 than small ones, while I have, at the same time, pointed 
 out how the disadvantages of the large rolls may be 
 overcome. 
 
vm 
 
 teanslator's preface. 
 
 I have, however, added nothing from American practice 
 to the book, because of the difl&culty which Herr Tunner 
 mentions, viz., that men who have invented anything 
 advantageous are naturally not inclined to allow their 
 experience to be made public in this way without some 
 advantage in return; therefore I have not mentioned 
 some important American modifications of rolling mill 
 practice, the foremost of which is the application of 
 the three high mill to rolling rails, and pecuKar arrange- 
 ments of the grooves of the rolls, which have been 
 made by the Messrs. Fritz, and adopted in most of our 
 largest mills. I have omitted these, however, with the 
 less reluctance since they are not absolutely essential 
 to a thorough knowledge of the practical principles of 
 roll-turning, and of the design of rolls, and because the 
 main object of this work is to inculcate these principles 
 carefully, and to illustrate them separately by a few 
 good drawings, rather than to offer a collection of ex- 
 amples of various kinds of work. I deem that this object 
 will be furthered by the fact that this work of Herr Tunner 
 offers a careful digest of European practice on almost 
 every point of the art, and believe that American mill 
 managers will find it profitable to study these rolls, designed 
 to work up various irons, many of them widely different 
 from our own, and compare them with those in use at 
 their own mills. 
 
 In respect to the nomenclature used, I have been obliged 
 to invent several names, in the choice of which I have 
 made it my object to prefer those names which already had 
 a practical meaning in any similar position — thus, for in- 
 stance, in the case of "JiUet" which term I have used to 
 designate the projections which separate Gothic and 
 
TEANSLATOK's PEErACE. 
 
 ix 
 
 similar passes, and whose surfaces are not at right 
 angles to each other, but may be at various angles 
 and of different forms. In the use of the word groove 
 I have not held it to be synonymous with pass, but to 
 represent only one-half of the latter, viz., the gToove 
 which, in any form of pass, must be turned upon one 
 or both rolls. In some varieties of passes this groove 
 is closed by a fillet on the other roll which projects 
 into it, and the surface of which is turned in such a 
 way as to give the desired form to the bar, generally 
 by means of strong pressure. I have, therefore, caUed 
 this fillet the ''former," a term which will, I trust, be 
 found acceptable, both on account of its practical sound, 
 and as clearly expressive of the uses of the part. 
 
 The measurements given in the text are in all cases 
 in Austrian inches and parts of inches, because the differ- 
 ence between these and the English is very small indeed. 
 Their exact equivalents may be found in the tables on 
 page 91 and succeeding pages. 
 
 All the figures in the atlas are to the scale of ^ natu- 
 ral size, unless otherwise specified. The dimensions are 
 marked ou them in Austrian inches, which may be con- 
 verted into English by means of the above tables. 
 
 I have thus endeavored to present to mill managers 
 and roll-turners, and all who take an interest in the art 
 of roll-turning, a carefully prepared system, by the aid 
 of which it may be possible to raise the practice to a 
 higher level. 
 
 John B. Pearse. 
 
ROLL-TURNI 
 
 Sec. 1. By tlie term roll, as used in tlie iron trade, is un- 
 derstood those cylindrical bodies whicli are made of cast- 
 iron — sometimes of cast-steel, and also, though seldom, of 
 wrought-iron. These cylinders are furnished with smooth 
 turned continuations at each end, and are accurately placed 
 and borne in a special frame or housing. A revolving motion 
 is communicated to the rolls through a shaft coupled to their 
 projecting ends, and this shaft may be driven by a steam en- 
 gine, or any form of water-wheel, either directly or indirectly, 
 by means of toothed wheels or belts. In each housing are 
 placed at least two rolls, which is the most usual number, 
 although the use of three together is becoming general, while 
 under special circumstances four have been placed in the 
 same housing. In the latter case, however, the rolls are 
 arranged in pairs, and the common axis of one pair need be 
 neither parallel to, nor in the same plane with, that of the 
 other pair. 
 
 That part of the roll which is between the smooth turned 
 continuations, and therefore lies free when in the housings, 
 is the part which is used for roUing the iron, and is called 
 the " hody " of the roll. The smooth continuation, which 
 revolves ia a special journal, supported in the housing, is 
 called the " neck,'" while the projecting ends upon which the 
 couplings take their hold, are called the "pods.'' 
 
 Fig. 1. 
 
 a 
 
 In the figure, a is the body of the roll ; 6 & are the turned 
 parts of the necks which rest in the journals or brasses, and 
 
2 
 
 c c are the pods wliicli project beyond the journals, and to 
 which either a shaft or another roll is coupled. Every 
 roll must have two necks. The notched part of the neck is 
 unnecessary when no further coupling is wished, and when 
 the upper roll is carried round merely by the friction occa- 
 sioned by the passage of the iron, both notched ends may be 
 dispensed with, but they are almost invariably added for 
 convenience' sake, as it might be necessary to use the rolls 
 coupled. Such rolls may be called " drag rolls" but differ 
 from the coupled rolls merely in not being directly driven. 
 The neck is not always cyhndrical in its whole length, but is 
 strengthened or curved outwards where it joins the body of 
 the roll, as is shown in Figure 1. This is especially the case 
 where the rolls must exert a great pressure on the piece to 
 be rolled. The coupled ends are often notched or star- 
 shaped, as in the figure, but may also be made square, or of 
 a circular section, with either two opposite or three equidis^ 
 tant notches, which are moderately deep. 
 
 For many purposes — such as rolling plate or polishing 
 hoop-iron — the body of the roll has the form of a smooth 
 cylinder. Such rolls are called plate or polishing rolls, and 
 are exemplified in Fig. 1. In other cases, the body of the 
 roll is made up of several cylinders, arranged like steps. 
 Such rolls ar§ called step rolls, and, in connection with pecu- 
 liar guides, are used in rolling flat iron, spring steel, etc. In 
 such cases they save many grooved rolls, which would be 
 otherwise necessary. Figure 33, on Plate III., represents 
 such a step roll, and Figure 34 shows the necessary guides ; 
 bothfi gures are size, and will be described hereafter. 
 
 § 2. By the term "pass " is understood those sections of 
 various forms which are produced by the relative position of 
 the different grooves and projections which are turned upon 
 the surface of the rolls ; the form of the pass appears sharply 
 marked on looking between the rolls when in position. The 
 term groove is sometimes used as synonymous with pass ; it 
 is, however, preferable to confine it strictly to the groove on 
 the body of a single roll. A pass may be formed by two cor- 
 
3 
 
 responding grooves, or by a single groove, into which there 
 fits a fillet or ring on the other roll. It is best to call this 
 ring the ''forrmr" in distinction to the term "collar,'' 
 which has reference, as generally used, rather to the fillet, 
 which, while dividing one pass from another, projects into a 
 corresponding groove on the body of the other roll. 
 
 Just as two rolls are necessary for rolling, so two corre- 
 sponding rolls are necessary to form a pass. It is, however, 
 very seldom that a pass is formed by more than two rolls, for 
 if three rolls are in the same housing (three high system), 
 they lie horizontal, parallel, and with their axes in the same 
 vertical plane, and the passes are formed by the middle, al- 
 ternately with the top and bottom roll ; therefore, by two 
 rolls only. It is only in isolated cases, such as the rolling of 
 thin iron tubes, that the pass is formed by four correspond- 
 ing rolls which work together. Therefore, in all following 
 descriptions, it must be always taken for granted that the 
 pass discussed is formed by two corresponding rolls, un- 
 less the contrary is expressly stated. Now, one roll almost 
 always lies vertically over the other, and in the same plane ; 
 therefore, in a two high train, where two rolls are used, they 
 are called respectively the hottom roll and the top roll. In a 
 three high train, however, when three rolls are placed above 
 each other in the same housmg, the third is called the middle 
 roll. 
 
 The passes formed by a pair of rolls are usually intended 
 for successive use, i. e., one directly after the other; but 
 other passes, or a set of passes, belonging to an entirely dif- 
 ferent series, may be turned on the same rolls, in which case 
 the two sets are entirely independent of each other. Be- 
 tween two neighboring passes there must be left a proper 
 space, which takes the form of a projecting rib or ring, and 
 is therefore called a collar. The two corresponding rings 
 at the end of each roll, which form the outside of the last 
 pass, are also called collars. The term collar, in its proper 
 sense, means a ring with rectangular edges. In order to dis- 
 tinguish those projections (other than formers, and not of 
 rectangular form) which separate the individual grooves of 
 
4 
 
 many varieties of rolls, it will be necessary to denominate 
 them hody-fillets and end-fillets respectively. The body-fillets, 
 so called, are those which are on the body proper of the roll, 
 and whose shape on either side is determined by the groove 
 on that side, while the end-fillets are those which lie at the 
 extreme ends of the roll, and which always have one rec- 
 tangular edge, i. e., the outer one, though the inner edge may 
 be of any form required by the adjacent groove. For the 
 purposes of these definitions it will be sufficient to consider 
 as collars all rings whose edges are rectangular, or nearly so, 
 while all those of which each side is obviously turned to form 
 one side of a groove, must be called body-fiUets. 
 
 § 3. The various forms of passes may be referred to the 
 following seven divisions, according to the position of the 
 pass and the way in which it is formed by the rolls. 
 
 1. Open Passes. — ^In this form, as shown in Fig. 2, part of 
 the pass is turned upon one roll, the 
 other and corresponding part upon 
 the other roll. The divisions be- 
 tween the passes are formed by a 
 fillet, turned on each roll ; there- 
 fore the pass (either the box pass a 
 or the Gothic pass b) is seen to be di- 
 vided between the top and bottom 
 roll, horizontally and nearly in the 
 middle, as are also the body-fillets, 
 c, d, e, and c', d' and e'. The body- 
 fillets remain, as it were, as rem- 
 nants of the original body of the 
 roll into which the necessary grooves have been turned ; 
 when the pass is large, however, it is roughly formed in cast- 
 ing the roll. The body-fillets of the top roll sometimes touch 
 those of the bottom roll when it is desirable to preserve the 
 size, and especially the height of the pass unchanged ; but, 
 as a rule, more or less space is left between them (a small 
 play is, for instance, observable in the drawing), whereby the 
 
6 
 
 pass is left more or less open at the point of division. There 
 are many and very different reasons why this play is left be- 
 tween the body-fillets. Sometimes the object is to save 
 passes, and then the top roll is raised for the first passage of 
 the bar, and lowered as desired for every succeeding pas- 
 sage ; but such a procedure should be resorted to only in 
 case of necessity. Again, it often happens that the hody- 
 fillets, or in other cases the collars, are kept apart, in order 
 that the size of the pass may be changed to the size which a 
 test-piece shows to be the proper one. This is the case in 
 finishing roUs, especially those for fine bar iron or rod, but 
 in these cases the play must be small. A third reason is, 
 that such play is a protection against breaking the fillets, to 
 secure which protection the construction described in Article 
 10 is also desirable- 
 
 2. Closed Passes. — In this form, as shown in Figure 3, a 
 groove is cut so deep into the body of the bottom roll that 
 the edge is higher than the total height of the pass. Such 
 Fig. 3. grooves are shown in the adja- 
 
 l -F m en do cent figure at a, 6 and c. Now, 
 
 in order to obtain the requisite 
 form in the pass, a projection, 
 viz., former, must be turned on 
 the upper roll, as at d, e, / (in- 
 stead of a groove, as in Fig. 
 2), and this former must fit into 
 the groove in the bottom roll, 
 thus closing the pass. The 
 whole pass is thus sunk into the 
 bottom roll, and the collars pro- 
 per, viz., g, h, i and h, are found 
 only on this roll, as they fit into corresponding grooves, I, m, n 
 and 0, in the top roll. By this construction the collars of 
 the bottom roll revolve with as little as possible side-play in 
 the grooves of the top roll, but a moderate space is left be- 
 tween the surface of the collar and the bottom of the corre- 
 sponding groove. This being the case, the height of the pass 
 
6 
 
 can be changed at will, as far as the vertical space allows, 
 by raising or lowering the top roll, without at any time open- 
 ing the pass. 
 
 It is evident that it is not absolutely necessary that the 
 groove should be in the lower roll, and the former in the 
 upper one ; but were the contrary the case, the introduction 
 of the bar into the pass would be rendered difficult, and con- 
 sequently the exertion required on the part of the rollers 
 would be much greater. The arrangement of grooves and 
 formers above described is, therefore, the usual one. "We 
 find, however, instances where the grooves are in the upper 
 and the formers in the bottom roll, as shown by Figures 59 
 and 65, on Plate V. The reason for this exception will be 
 given in Article 26, in which the construction of rolls for 
 angle iron is discussed. 
 
 EoUs mth closed passes cost more than those with open 
 ones, but possess the important advantage, that the sides of 
 the pass form a certain guide for the bar at its entrance and 
 exit. 
 
 3. Half Open Passes. — ^These are passes which are closed 
 on one side and open on the other, at the pitch line. This 
 form occurs occasionally where difficult and complicated sec- 
 tions are to be rolled, and is illustrated in Fig. 40, on Plate 
 III., by the last two passes on the right. The case also oc- 
 curs where the immediate bounding lines of a pass would 
 seem to indicate an open pass, which is, notwithstanding, 
 seen to be a closed pass on regarding the remoter bounding 
 lines. An instance of this latter kind occurs on Plate V., in 
 the first two passes (on the left) of Fig. 61. As the above 
 cases have the main characteristics of a closed pass, they are 
 commonly regarded as such, and are not specially classified. 
 
 4. Edging Passes. — The position of these passes is gener- 
 ally sufficient to distinguish them, but their main character- 
 istics appear distinctly only during use, as their function is 
 to exercise a powerful pressure on the bar, and because the 
 bar is turned only 90 degrees before its introduction into 
 
7 
 
 sucla a pass, instead of 180 degrees, as in the use of all others 
 — that is, all others on a single pair of rolls, as in a three 
 high train, nearly finished bars not turned over when they 
 have been brought nearly to the required form. The edging 
 passes are generally closed passes, as is shown in Fig. 40, on 
 Plate III., by the first pass on the left ; or on Plate IV., in 
 Fig. 44, by the last pass on the right ; in Fig. 45, by the 
 middle pass ; and in Fig. 46, by the first pass on the left ; 
 or on Plate VI., in Fig. 68, by the last pass on the right ; 
 and in Fig. 71, by the pass on the right. These passes may, 
 however, be open, as is shown by the last pass on the right 
 in Fig. 16, on Plate II., and also by the last pass on the right 
 of Fig. 53, on Plate V. 
 
 The so-called adjusting passes, as well as the flatting passes, 
 are named from the nature of the work they do ; they re- 
 semble the edging passes in their position, and their names 
 are, therefore, mentioned here, while their special description 
 will be presented hereafter. 
 
 5. EccENTEic Passes. — These are, as the name impKes, 
 passes which are turned eccentrically to the axis of the roll. 
 As a rule, only one roll, and generally the bottom one, is 
 
 Fig. 4. 
 
 turned eccentrically, as shown by Fig. 4, which is a section 
 of a pass for rolling the so-called fish-belly rails. Sometimes 
 
 2 
 
8 
 
 tli^t6p roiy^.tiirned eccentrically, but it is very seldom that 
 botli aife ^©''turned. 
 
 The 'fcar which has gone through such a pass is naturally 
 of varying depth, and the point of least depth repeats itself 
 at intervals corresponding to the circumference of the bottom 
 of the pass. 
 
 It is evident that no more than one such eccentric pass 
 can be used on the same bar, and also that it must be the 
 last. Such passes occasion trouble in rolling, and increase 
 the percentage of scrap and imperfect bars ; therefore, they 
 are seldom used.* 
 
 6. Spiral Passes are those in which the bottom of the pass 
 is not a circle but a spiral, which, more or less, nearly ap- 
 proaches a circle, and in which the under line is joined to the 
 
 Fig. 5. 
 
 Tipper, just before its passage under the latter, by an arc of a 
 circle, as represented in Fig. 5. 
 
 * The best method of avoiding such a pass is to so arrange the rolls that 
 rtheir relative position in a vertical plane may be changed during the passage 
 -of the bar. Such an arrangement is not similar to that of a tyre mill, as in 
 the latter the rolls can be moved at vnll nearer to each other ; here, however, 
 they are gradually moved to and from each other at a rate and to an extent 
 which are definitely fixed beforehand. 
 
The upper roll is the one usually turned in t&i«^i^^ l^i^'' 
 form resembles the eccentric, but has the peculi^tit^^Ht 
 gives a much more gradual taper to the bar than the latter. 
 This is, therefore, the form of pass applied in the small 
 rolls used for pointing-wire, and in the larger ones which 
 give the wedge shape required by bars for some varieties of 
 springs. The end to be rolled out into a wedge is thrust 
 quickly through the pass as soon as the depression appears, 
 and is then rolled out again directly back into the hands of 
 the roller. A stop on the other side of the rolls prevents the 
 bar from being thrust too far, and it is usual to turn several 
 of these passes of various sizes in the same roUs, in order to 
 be able to taper out finished bars of various shapes and 
 sizes. In order to produce this form of groove, it is neces- 
 sary to turn an eccentric groove, and then chip off as much 
 metal from its highest side as is required to form the desired 
 spiral. 
 
 7. Intermittent Passes.— The peculiar feature of these 
 passes is that the bottom of the pass is made either with 
 notches or projections of various forms, which are applied to 
 either roll smgly, or to both, as made necessary by the 
 finished form desired. The notches or projections must, of 
 course, be placed at an equal distance from each other, in 
 order to avoid waste. Such a pass is represented in Fig. 101, 
 Plate X., as apphed to the production of " spike-rod." Like 
 the eccentric passes, these can only be used to finish. By 
 varying the form and position of the notches or projections, 
 a very great number of different forms may be produced. 
 
 § 4. Another and a very common classification is made ac- 
 cordmg to the shape of the pass. The following seven va- 
 rieties are usually distinguished : 
 
 1. Gothic Passes.— These passes have the form of a 
 Gothic arch, as represented in Figs. 1 to 7 inclusive, on Plate 
 I. They are always open passes. 
 
 2. Flat Passes (to which class the Box Pass belongs).— 
 The form of these passes is a rectangular parallelogram of less 
 
10 
 
 height than width, as shown in Figs. 8 to 12, on Plate I., and 
 in Figs, 16 to 29, on Plate II. A flat pass is always a dosed 
 one / when made an open one, it becomes a hox pass. 
 
 3. Diamond Passes. — These are square in their general 
 form, and are represented in Figs. 13 and 14, on Plate I., or 
 better in Fig. 30, on Plate II. They are, with very few ex- 
 ceptions, open passes. 
 
 4. Oval Passes. — These are shown in Fig. 35, on Plate 
 III., along with diamond passes ; they are, like the latter, 
 almost always open passes. 
 
 5. Round Passes. — These are similar to those drawn on 
 Plate III., in the last pass of Fig. 35, and in the right half 
 of Fig. 36. They are always constructed as open passes. 
 
 6. Polygon Passes. — In this division are included prin- 
 cipally hexagon and octagon passes. The latter form may 
 be found in the left half of Fig. 36, on Plate III. These are 
 always open passes, and are often included in the following 
 class. 
 
 7. Shapes. Under this head are included all those passes 
 which are differently formed from any of the preceding, and 
 among them there exist very complicated forms. As a rule, 
 these passes are of the closed form. 
 
 § 5. In addition to the two methods of classifying passes, 
 as above described, they may be individually named accord- 
 ing to the work which they are to perform. Considered in 
 this light, the four following kinds of passes may be dis- 
 tinguished : 
 
 1. Welding Passes. — By this term is understood those 
 passes through which the piece is rolled while at a welding 
 heat. These are naturally the first passes, and generally the 
 first three, and are made with a very great draw or reduction, 
 in order that the pile may be powerfully compressed. Their 
 surfaces are often roughened with the chisel, in order that 
 they may take firm hold of the pile. The forms most gener- 
 
11 
 
 ally given them are the Gothic or flat, with the exception 
 that, in the case of a good many large shapes, the form of 
 the first pass must be accommodated to that in which it is 
 necessary to make the pile. 
 
 2. Deawing Passes. — In these the section of the pile is 
 often not at all regarded, the main object being to reduce it 
 as fast as possible. This character is especially marked in 
 rolls for merchant bars. In rolls for rails, I beams, etc., the 
 final form cannot be disregarded ; hence in these, and in all 
 similar cases, the rapid drawing down of the pile or bar 
 must be accompanied, to some extent, by a shaping process. 
 These passes are used immediately after the welding passes, 
 and are often similar to them in form. The Gothic form is, 
 however, the most useful, as it admits of a very heavy draw, 
 as will be hereafter explained. Any number of these passes, 
 up to 10 or 12, may be used successively in one heat. Both 
 the above classes are often collectively included in first and 
 second roughing passes, and may be turned together on one 
 pair of rolls, though often constructed on two. 
 
 3. SHAPiNa Passes. — These mould the bar gradually to the 
 desired form, at the same time reducing its sectional area. 
 In the first of these passes the reduction is great, while the 
 bar remains hot and soft, and, consequently, flatting passes 
 are often used to increase the reduction, as well as to assist 
 in shaping. The number of these shaping passes varies 
 greatly. In the case of the simpler forms they are constructed 
 as the last preparatory or roughing passes, but when compH- 
 cated shapes are rolled they often form the necessary welding 
 passes. Such a case is presented in Fig. 82, on Plate VIII., 
 which delineates the passes for large deck beams, or T iron. 
 
 4. Finishing Passes. — The object of these passes, which 
 are the last in the train, is to complete the form of the bar, 
 and make its surface smooth and clean. The reduction 
 given these passes is very small, and the contraction of the 
 finished bar on cooling must be provided for in the last one. 
 Strictly speaking, the last pass only could be called a finish- 
 
12 
 
 ing pass, but the piece is often rolled several times througli 
 it, as in the case of the heavier round iron. Strictly speaking, 
 the last pass only could be called a finishing pass, but the 
 piece is often rolled several times through it as in the 
 case of the heavier round iron. Sometimes two finishing 
 passes are placed side by side, or a rough space is left 
 for the second, in order that the latter may be ready for 
 use, or easily turned up, when the first is worn out. Some 
 of these passes always precede the finishing pass on the 
 same roll. When the terms first and second roughing are 
 used, most of these passes are included with the finishing 
 passes proper, under the term of finishing passes. 
 
 The following varieties may also be named from the 
 nature of their service : 
 
 5. Flatting Passes. — These are distinguished from all 
 others in that they greatly increase the width of some part 
 of the bar rolled through them. To this end, they must 
 exert a considerable draw, and in this respect, as well as in 
 their position, they are ranged with the edging passes. They 
 are extensively used in rolling shapes, such, for instance, as 
 T rails. 
 
 6. Adjusting Passes. — These are used, when the finishing 
 is accomplished with step or pohshing rolls, in order to give 
 the bar the desired width, and finish its corners. To ac- 
 complish this end, the bar must be rolled on its edge, and 
 the pass, therefore, must be made high and narrow. If the 
 bar is to be finished with rectangular corners, the use of a 
 special guide arrangement, as shown in Fig. 34, on Plate III., 
 is often preferred to that of the adjusting pass ; but the use 
 of the latter cannot be well avoided if the comers of the 
 piece are rounded off. These passes are, with few excep- 
 tions, open. 
 
 § 6. If we regard these respective divisions more gen- 
 erally, we find that the six kinds of passes described in § 5 
 separate themselves into roughing and finishing passes. 
 This is especially the case in ordinary bar-iron rolls, and 
 these designations are quite convenient. This being the 
 case, the rolls which contain these respective passes can be 
 
13 
 
 respectively called roughing and finishing rolls, which are 
 the practical designations. When there are two sets of 
 roughing rolls, those which contain the welding and part of 
 the drawing passes are called the first roughing rolls, while 
 those which contain the other drawing and part of the 
 shaping passes are called the second roughing rolls, and 
 those containing the rest of the shaping and the finishing 
 passes are called the finishing rolls. These terms agree with 
 those used in forging, viz., ih.e> finishing or polishing the forg- 
 ing which has been previously drawn down out of a larger 
 pile or ingot. 
 
 As we have stated above, the passes necessary to the 
 production of any desired shape need not be confined to two 
 rolls only. They are sometimes all turned on a single pair, 
 but much more generally divided among three, or sometimes 
 five, or even a greater number of pairs. The passes are thus 
 divided, not only to avoid too great a length between the 
 necks, but also to get more room for the men by separating 
 the pairs. Sometimes it is an object to be able to alter at 
 will the height of certain passes, particularly the last two 
 finishing, or to get a diflferent angular velocity {i. e., surface 
 speed) for special passes, which object is attained by group- 
 ing them as far as possible on different rolls. In rolling 
 wire rod, it is necessary that the bar should be in several 
 passes at a time, and it is therefore desirable to so separate 
 these passes as that they may be conveniently reached and 
 worked. 
 
 It is, also, not necessary that a pair of rolls should be 
 turned exclusively for one section, but there may be upon 
 the rolls passes of widely different sizes of the same, or even 
 of an entirely different kind. 
 
 When a number of rolls, with their respective housings, are 
 set up in a continuous line and so coupled that their indivi- 
 dual velocity must be the same, the combination of rolls and 
 housings is called a train of rolls. The various trains are 
 named according to the kinds of iron they produce. Thus a 
 train which receives squeezed or hammered balls and trans- 
 forms them into puddle bar, is called a puddle train, while a 
 series of rolls for rails is similarly called a rail train. The 
 
14 
 
 trains devoted to round or square iron, small bars of fancy 
 sections and angle irons, are, in a general way, called mer- 
 chant trains, wliicli, in a large mill,, may include heavy and 
 light bar, or wire-rod trains, etc. 
 
 § 7. After the foregoing general view and classification of 
 the passes which commonly occur in a set of rolls, it will 
 be most useful to consider the way in which the individual 
 passes act upon the iron, before their construction in detail 
 is discussed. 
 
 A pass, no matter what its form, can exert a direct pressure 
 on the iron only in a perpendicular direction, at right angles to 
 the axis of the roll. This pressure diminishes the height of 
 the bar rolled, but does not materially affect its width. Now, 
 as this reduction of height occasions a corresponding increase 
 of length, we may call this pressure the draw pressure. The 
 difference in section between successive passes is called the 
 draw or draught, and is occasioned mainly by the difference 
 in height of the respective passes. It is clear that the width 
 of the bar cannot be directly affected, since the sides of the 
 pass must be parallel in a horizontal (though not necessarily 
 in a vertical) direction, in order to permit the passage of the 
 bar. Therefore, the sides of a pass can exercise no direct 
 pressure on a bar, and cannot take hold of one wider than 
 their own distance apart. Such a bar can scarcely be intro- 
 duced, and the excessive iron is forced into all the interstices 
 of the pass, forming fins, which are torn off, or remain to 
 materially injure the surface of the bar. 
 
 In order to facilitate the rolHng, it is necessary to construct 
 each of the successive passes a little wider than the prece- 
 ding. This difference in width varies fi-om almost nothing to 
 an eighth of an inch or more, according to the size and form 
 of the passes. It follows, therefore, that the successive pass- 
 es should be made wider and wider, and this is indeed com- 
 monly the case when the bar is either not turned over at all, 
 or turned 180 degrees after each pass. It would, however, 
 be extremely inconvenient to thus widen any large number 
 of successive passes, and it becomes necessary to intersperse 
 
15 
 
 flatting passes (vid. § 3) among the rest ; for these passes the 
 bar is turned only one quarter over {i. e. 90"), whereby the 
 former thickness or depth of the bar becomes its width, and 
 the former width becomes its depth. The pass succeeding 
 the flatting pass is then narrower than the one which pre- 
 ceded the same, and forms the starting-point of a new 
 series, which is again intercepted by a flatting pass, when the 
 width of the grooves becomes inconveniently great. There 
 are, hovever, some forms of passes, such as the Gothic and 
 Diamond, in which the bar must be turned quarter over at 
 each pass, and, consequently, the depth and width of the bar 
 interchange positions at each pass. Here the successive 
 passes must be so proportioned that the height or depth of 
 one pass is a little smaller than the width of the next. This 
 arrangement of passes is so important that it is termed by 
 Truran, " the fundamental principle of rolling." 
 
 In flat passes, however, and in some shapes, the necessity 
 of increasing the width of the successive grooves may be par- 
 tially and even wholly avoided, by slightly increasing the width 
 of each groove from the bottom of the same, outwards. This 
 construction enables the pass to take in easily a bar which 
 otherwise could scarcely be introduced, and is of additional 
 use, in that it greatly facilitates the exit of the bar. 
 
 Although the vertical draw-pressure is the only one direct- 
 ly exercised by the grooves, yet a side-pressure is indirectly 
 occasioned, since the iron is more or less soft and yielding, 
 and is, therefore, not only drawn out lengthwise, but also 
 bulged out sidewise and pressed against the sides of the 
 pass, if the latter is not wide enough to allow the pile or bar 
 to spread without contact. This is seldom the case, and 
 hence the sides, in preventing further spreading, exercise a 
 pressure opposed to the draw-pressure. Let us, therefore, 
 call the former force the side-pressure. Now, this side-pres- 
 sure will be gi-eater as the draw is greater, the width of the 
 pass remainiag the same, and vice versa, the greater the 
 width the less the side-pressure, the draw remaining the 
 same. The side-pressure must be correctly proportioned to 
 the draw or draw-pressure, in order that the bar may take 
 
16 
 
 the exact form of the pass ; when the former is too great, 
 the iron is forced into the interstices between the body-fillets, 
 or between the collars and the formers, which necessarily 
 cannot touch each other, and a fin is formed as well in the 
 case of closed as of open passes. Therefore, when the piece 
 must be quickly reduced, as in drawing or shaping passes, a 
 powerful draw must be employed, and, therefore, the widths 
 of all the passes should be made proportionately greater, on 
 account of their considerable width as related to their depth. 
 For instance, the oval passes are extensively used in reducing 
 small sections. 
 
 This relation between draw and side pressure is obvious, 
 and, therefore, easily understood. It is, however, not so 
 clear why a piece of iron which is at a higher heat, or which, 
 on account of its chemical or physical character, is softer 
 than another, widens proportionately less, and hence occa- 
 sions less side-pressure than the latter piece, which is harder, 
 and resists the draw more strongly. The following experi- 
 ment proves, however, that such is actually the case : Four 
 bars, two of which were of soft iron and two of steel, were 
 rolled through the same closed pass into flat bars 1 inch wide 
 and I of an inch thick. An iron and a steel bar were brought 
 to a cherry red heat, while the second iron and the second 
 steel bar were brought nearly to a welding heat. All were 
 then passed through polishing rolls, furnished with accurate 
 guides, and reduced at one passage to a thickness of of 
 an inch. In this way aU were exposed to the same draw- 
 pressure, while entirely at liberty to expand laterally. The 
 result was, that, in the case of the iron, that bar which had 
 been rolled at the lowest heat was several per cent, wider 
 than that rolled at the high heat, and the same was true in 
 the case of the steel bars ; it also appeared that the steel 
 bar was in each case somewhat wider than the iron one which 
 had been rolled at the same heat. The latter result was 
 especially surprising, but its correctness cannot be doubted ; 
 its accuracy is further proved by the fact that it is generally 
 known among merchant iron rollers that in rolKng long hoop 
 iron the rear end of the bar, which grows comparatively cool 
 
IT 
 
 before it is rolled, is always appreciably wider than the other 
 end, even though it remains somewhat thicker. 
 
 It is easier to understand the fact that if the same pass is 
 turned on rolls of different diameters the iron will be length- 
 ened more, and widened less, in the roUs of small than in 
 those of large diameter. The same principle applies here as 
 in the working of a hammer with broad or narrow die : the 
 narrower the die the quicker is the piece drawn out ; and if 
 it is desired to forge a thin plate under a broad die it must 
 be forged out crosswise, and with frequent turning under a 
 special narrow tool, which thereby supplies the place of a 
 narrow die. 
 
 § 8. At first sight it appears that it would be the best and 
 most natural arrangement to turn one half of a pass on the 
 top, and the other on the bottom roll. In this case the haK 
 of the depth (sometimes the half of the sectional area) of the 
 pass would be under, and the other half over, a horizontal 
 line drawn midway between the axes of the rolls, which line 
 is called the 'pitch line. If the pass were thus divided any 
 force would easily turn the bar, at its exit, either up around 
 the top roll or down around the bottom roll. This some- 
 times happens to the slabs of a large pUe, when the bottom 
 or top ones are torn off and curled up. Now, if a bar, or a 
 single slab of a pile happened so to curl up, it would natu- 
 rally not only make the same worthless, but also be likely to 
 break a roll, or occasion other damage, such as destroying 
 pipes or guides, etc., and, therefore, special appliances must 
 be used to prevent these accidents. But if the depth of the 
 pass was equal above and below the-pitch line, these appli- 
 ances would be necessary on both roUs — an arrangement 
 which would not only be difficult, but also inconvenient; 
 therefore the depth of a pass is always so divided that a 
 little more lies under the pitch line than above it, when the 
 pass is designed for ordinary two high rolls. In other words, 
 the diameter of the groove on the top roll is made somewhat 
 greater than that of the other groove on the bottom roll. 
 The upper surface of the bar is, therefore, lengthened or 
 drawn out more than the under one, which causes a tendency 
 
18 
 
 in the bar to curl down toward the bottom roll. Thus it be- 
 comes possible to use only one set of appliances, or guards, 
 on the under roll, as there is now no danger that the bar will 
 curl around the top roll. 
 
 There are, further, two kinds of these appliances ; the one 
 is horizontal and projects into the pass, and is called a guard ; 
 the other kind stands vertically behind the pass, and on each 
 side of it, so as to prevent the bar from swerving to the side. 
 These upright pieces are termed guides, and are supported 
 on a cast-iron plate called the " hearing plate,'' the front edge 
 of which sometimes takes the place of a special guard, as 
 detailed below. 
 
 The guards consist, in their simplest form, merely in a 
 cast-iron plate, which is laid behind the roUs, and so formed 
 that it projects into each pass, though not accurately fitted 
 to the same, thus loosening bars from the rolls if they have a 
 tendency to curl down as they come out. This cast-iron 
 plate, or " hearing plate" also carries the guides, as men- 
 tioned above. When, however, the section of the pass is 
 small, or its form complicated — in which case it would be 
 difficult to loosen the bar — it is necessary to lay upon this 
 cast-iron plate a flat wrought-iron bar which has the exact 
 shape of the pass, and closely fits the curve of the roll. If 
 the pass is irregular — as an eccentric pass, for instance — it is 
 best to place a second guard under the first, slanting up 
 against the roll from a bar placed under the bearing plate. 
 This guard is kept by its own weight against the bottom of 
 the groove. 
 
 If the difference between the diameters of the working 
 surfaces of each groove is made too great, the guards are 
 unnecessarily strained and rendered useless, while the bar 
 itself is injured by the unequal draw of its surfaces. The 
 proper difference of the diameters varies with the character 
 of the iron and the circumstances of its treatment, and must 
 be carefully ascertained and regulated. The following rules 
 represent the general practice : 
 
 In the case of plain open passes the diameter of the top 
 roll is from to g of an inch greater than that of the bot- 
 
torn roll. In the case of closed flat passes the groove in the 
 bottom roll is cut so deep that its diameter is from to | of 
 an inch less than that of the former which closes it. In rolling 
 shapes it sometimes happens that more than § of the pass is 
 under the pitch line. In the case of larger roughing passes 
 and of flatting or edging passes, the differences of diameter 
 of the working surfaces vary from ^ of an inch to 1 inch, 
 and even more. 
 
 In all the drawings of grooved rolls, or of passes, which 
 vsdll be given hereafter, the diameter of the working surface 
 of the grooves will be added in figures, or the pitch line of 
 the pass will be drawn in. The classification found most con- 
 venient in treating the construction of the various passes in 
 detail is the one described in § 4. It has been my aim to 
 discuss in the following pages the construction and the draw 
 of series of passes, to treat of their application, and, as far 
 as possible, illustrate all descriptions of roUs by means of 
 accurate (working) drawings to the scale of appending 
 the respective passes in full size. 
 
 § 9. The Gothic passes are very important, are frequently 
 used, especially as rougliing passes, and are constructed and 
 apphed in many different forms. They possess the advantage 
 of being simple and durable, while they do not chill the bar 
 irregularly, as their form approaches a circular one, and, on 
 the other hand, as their form is so nearly square they draw 
 the bar equally on all sides. This form was described by 
 Karsteii in the last edition of his 
 work on iron, in Figs. 1 and 2, on 
 Plate LIII. ; but the construc- 
 tion there given is entirely incor- 
 rect, as the depth of the pass is 
 much greater than its width, and, 
 so far as my experience goes, this 
 form has never been used. 
 
 A very simple and practical 
 construction is shown in Fig. 6, 
 the depth and width of the pass 
 being supposed to be given. 
 
 Fig. 6. 
 
20 
 
 To construct the pass, draw the straight lines a h and c d at 
 right angles to one another ; lay out from their intersection 
 at 0 the half of the given width on each side in a and h, 
 as well as the half of the given height above and below 
 the horizontal line in c and d. Then with the radius a h 
 describe from the point h an arc near the point m. With 
 the same radius describe an arc from d which intersects 
 the former at m. Further, with the same radius describe 
 from the poiut m the arc b d, which is one side of the pass. 
 By the same process repeated from each point, a c and d in 
 turn, the points n p and q are found, and the respective arcs 
 described, as shown in the figure. The sharp comers at a 
 and b must be rounded o£f, as shown in § 10, Fig. 9. 
 
 In this construction the first point to be considered is, 
 whence the given depth and width are derived. Now, the 
 absolute dimensions of the first pass are regulated by the 
 size of the pUe to be rolled, or, vice versa, the size of the 
 piles must be regulated by that of the first pass in the 
 roughing rolls on hand. In most cases it suffices to give the 
 first Gothic pass such dimensions that a pile of 5 1 to 6 inches 
 high can be taken in. This is the practice for puddle blooms 
 and small pUes, but for larger piles the Gothic pass is seldom 
 used. The only case, in my experience, in which this form 
 was used with more than 7 inches average diameter, was in 
 an EngUsh mill, with the special purpose of taking in the 
 puddle balls direct from the furnace, in case the forge ham- 
 mer should suddenly break down. In order to have no 
 difficulty in rolling the balls, they were made rather cylin- 
 drical than spherical. It generally suffices to give the first 
 pass an average diameter of only 5 inches, as when piles 
 larger than this are rolled, the rolls can be opened for the 
 first two or three passes by raising up the top roll for, and 
 lowering it after each passage of the pile. 
 
 Further, the usual difference between the depth and width 
 of a pass is mostly about one-sixth of the average diameter 
 of the same, hence about one inch for the first and largest 
 passes. Therefore the depth and width of the first pass may 
 from these data be easUy suppHed in special cases. 
 
21 
 
 Another method of construction is given in Fig. 7, which 
 is extensively used in Upper Silesia and Austria, and origi- 
 nated with the Nestor of Continental rollers, Mr. Talbot. In 
 this method, a circle of 
 the average diameter de- 
 sired is used as the basis, 1^ 
 and is hence called the '^^^^'T^^.^ 
 Construction circle. In 
 this construction circle 
 the horizontal and verti- 
 cal diameters a b and c d 
 are drawn and continued ^ , 
 
 on each side beyond the ^"Har ■^V' 
 
 circumference. The hori- 
 zontal diameter is divided into five equal parts, and one of 
 these parts is subdivided into eight. The lines a a' and hb' 
 are laid off of a length equal to five of these subdivisions, 
 and the points a' and &' denote the extreme horizontal 
 width of the pass. With a radius of ^ of the diameter a b, 
 lay off the points efg and h from the points a and b ; then, 
 with a radius of 2| fifths of a 6 {i. e., |i of the diameter), de- 
 scribe arcs from a' and b', and mark the points of intersec- 
 tion at m n, m' and n'. These four points are the centres 
 of the required arcs, hd,df, etc., which form the sides of the 
 pass, and are described from these points with the radius 
 mhorn/, etc. From the points g and b^ describe, with a 
 radius equal to | of the diameter, two arcs cutting one 
 another in the point r, and from this point, with the same 
 radius, describe the arc g b' ; repeat the operation for each 
 arc b' Ti, e a', and a'/, and the pass will be properly closed. 
 
 The constructions illustrated in Figs. 6 and 7 * are chiefly 
 used for puddle roughing roUs, but for the roughing roUs of 
 a merchant or guide train, the followiag construction is pre- 
 
 * It is not possible to explain either these or any following forms of passes 
 on any mathematical or physical principles ; but the various constructions 
 given represent forms which have been gradually determined by experience, 
 while a fixed and approved measure is given for each curve and line which 
 forms any given pass. 
 
22 
 
 iev4^- ,4^1iis is;§ii0wn in Fig. 8, and differs from that shown 
 m e points e and / and g and h are joined 
 
 respeetij?-wy«^o a' and b', not by means of arcs of a circle, but 
 by straight lines, and that the extension of the horizontal 
 diameter of the construction circle, and consequent width of 
 
 the pass, is somewhat less. The 
 ^' horizontal diameter a & is divided 
 
 as before into 5 equal parts, but 
 one of these is subdivided into 
 16 parts, and the lines a a' and 
 h h' are laid off with a length 
 equal to 8 of these subdivisions. 
 The four points, m w, n and n^^ 
 are are found as before, from 
 a' and b', but with a radius of 
 2x^7 fifths of a & (i. e., |^ of the 
 diameter) ; the arcs which form the sides are then further 
 described, as in Fig. 7. The form of Fig. 8, as compared 
 with that of Fig, 7, shows a little less width and a greater 
 height ; the pass is, therefore, better closed at the sides, a 
 form which is desirable for the more compact material (i. e., 
 pile of flat bars) to be rolled, which contains by far less 
 slag than puddle balls, and out of which the slag is rolled 
 far more easily than out of the latter, because it is hotter. 
 
 For the roughing rolls of a bar train in which the draw 
 (or draw pressure) must be very great, in order to reduce as 
 quickly as possible, when for this reason, the use of oval 
 passes does not seem desirable, it is advisable to draw the 
 arcs described from a' and b', and intersecting the circle at 
 m n, m' and n' with a radius of 2x*V fifths {i. e., If of the 
 diameter), whereby the depth of the pass is slightly re- 
 duced. 
 
 § 10. The draw or draught of a pass is, as explained in 
 § 7, a difference between its area and that of the next larger 
 pass ; and in the case of the Gothic form, as well as in that 
 of some other forms, the width of one pass must be a trifle 
 in excess of the height of that which precedes it. The dif- 
 ference varies with the absolute size of the pass, amounting 
 
23 
 
 in small passes to about of an inch, an^^We j%^e| to 
 about i of an incb. A greater excess is S^[^ul£^be 
 first two passes of puddle roughing roUs, as it is necessary 
 that these should have ample width in order to safely take in 
 blooms which are irregularly formed. In these passes the 
 width of one is often from i to § of an inch greater than the 
 height of the preceding. The draw must also be regulated 
 by the quality of the iron to be rolled. A good quality bears 
 without injury a strong pressure, and rolls smoothly under a 
 draw which would tear a poor iron to pieces. It is, however, 
 scarcely necessary to remark that an unnecessarily Kght 
 draw delays the rolling greatly, and, therefore, iacreases the 
 amount and cost of the required labor. The draw should 
 obviously be as heavy as possible, but local experience only 
 can guide to a correct construction. Hence this local expe- 
 rience is exceedingly valuable, but is too often inaccurate, 
 and even imaginary. 
 
 The draught of Gothic passes is on the average about yV, 
 but increases with the size of the passes up to I or |. Small 
 passes, with a draw of |, would be useless, as such a reduc- 
 tion would cause the formation of fins— thin, riband-like 
 longitudinal projections along the sides of the bar, where it 
 has been forced against and between the body-fillets. With 
 reference to the diameter of the construction circle, a simple 
 and approved construction for the Gothic passes of puddle 
 rolls, which is used at many works, is the following : 6 inches 
 5, 4i 3|, ^, 2i, 21, 2f, 2f, 2, and 1| inches diameter of the 
 circles of the successive passes. Here we see immediately 
 that more attention has been paid to simplicity than to 
 accuracy. The first, or sometimes even the second, of these 
 passes is skipped when the balls have been made too small, 
 whUe the smaller passes are only used as far as the kind of 
 puddle bar, which is to be produced, requires. For instance, 
 it often happens that the five passes, from that of which 
 the construction circle is 5 inches diameter to that of 
 which it is 2| inches diameter, are the only ones used when 
 the balls are formed in the furnace to a weight of about 1 
 cwt., and rolled into bars 3 inches wide. The above reduc- 
 
 3 
 
24 
 
 tion, or draught, I would designate as an average one, as it 
 is often increased for a good quality of puddled iron, and as 
 often diminished for a poor quality. 
 
 The body-fillets of a roll he between the individual passes. 
 If they are too narrow their edges are easily broken off ; if 
 too wide, the rolls are too long, or, in other words, fewer 
 grooves can be turned upon an equally long roll. There is 
 another reason why narrow fillets are preferred by many, 
 viz., that in case any tool — the tongs, for instance — is carried 
 into the rolls, it is easily broken up by the narrow rings 
 without breaking a roll. As is natural, tlie absolute width of 
 the body-fillets increases with the size of the passes and of 
 the rolls, but not by any means in a direct proportion. In 
 the case of small rolls the width of the body-fillets is usually 
 from ^ to I of an inch, while it often amounts to an inch or 
 more on larger rolls. The end-fillets are always made a little 
 wider than the body-fillets, in order to obtain extra strength, 
 and on account of the fact that the roUing would be difficult 
 if the pass was brought too near the housing, as would be 
 the case if the end-fillets were narrow. 
 
 If the body-fillets revolve closely on each other, their cor- 
 ners soon become ragged, as they lie nearest the pass, and 
 are therefore more strongly heated and expanded than the 
 rest of the ring. 
 
 This evil is remedied by rounding 
 off the corners in. the lathe, as shown 
 in Fig. 9, even when the construc- 
 tion of the pass does not require it. 
 As a rule, the construction of the 
 pass does require that the comers of 
 the rings be rounded off (as in Fig. 
 7, on page 21), with the object of 
 preventing the permanent formation 
 of fins by making a thick one, which, 
 at the next pass and consequent 
 turn of the bar, will be thoroughly rolled in again. Another 
 very common precaution of this kind is to make the inner 
 comers of a pass of such shape that the metal can bulge out 
 
25 
 
 considerably at the next turn without being forced into the 
 interstices of the rolls. 
 
 In many cases, especially on puddle roughing rolls, the 
 body-fillets are not allowed 
 to touch at all, and are be- 
 sides rounded off over their 
 whole width, as represented 
 by Fig. 10, in full size. 
 Here the name open pass is 
 seen to be very characteris- 
 tic. Such open passes allow 
 the cold, stiflf puddle slag to 
 find its way freely out of the 
 ball. Other reasons why the 
 rings should not touch, are adduced in § 3, in the general 
 description of open passes. 
 
 § 11. On Plate I., in Fig. 1, is shown a pair of puddle rolls, 
 used in a Westphalian mill, viz., that at Horst (scale full 
 size), and in Fig. 2, the corresponding passes are given in 
 full size. The construction of these passes is that detailed 
 on page 20, and shown in Fig. 6, with the rounded corners 
 described in the preceding paragraph. There are also shown 
 on Plate I., in Fig. 3, the puddle rolls of a Styrian mill, in ^ 
 full size, with their passes in full size in Fig. 4. The con- 
 struction of these last is that explained in Fig. 7 of § 9. 
 
 A comparison of the two pair of rolls shows that the 
 draught of the Styrian rolls is rather lighter than that of the 
 Westphalian, although the great superiority of the Styrian 
 iron would lead one to suppose that the contrary would be 
 the case. But the Styrian rolls must also be used to roll 
 hard iron (fine-grained iron), and even puddled steel. Fur- 
 ther, the Styrian balls are, as a rule, very thoroughly ham- 
 mered, and must thus be rolled when pretty cold and quite 
 hard and solid, even when the iron itself is very soft. — EoUs 
 which contain both Gothic and flat passes will be discussed 
 under the head of those containing only flat passes. 
 
 Inasmuch as Gothic passes are habitually used as welding 
 
26 
 
 and drawing passes for bar-iron rolls, even of quite small 
 size, therefore Fig. 5, on Plate I. is added ; this represents 
 the roughing rolls of a bar-iron train, in full size, while 
 the corresponding passes are shown in full size in Fig. 6. 
 These passes are drawn according to the method described 
 ii § 9, Fig. 8. Figure 7, on Plate I., represents in full size 
 merely the roughing passes of a train for small bar-iron, 
 which are constructed similarly to those of the puddle rolls 
 shown in Figs 1 and 2, on Plate I. Both of these rolls, viz.. 
 Figs. 5 and 7, are Styrian, and the latter is in use at Neu- 
 berg. 
 
 As may be seen in the drawings, the draw of these passes 
 is about the same as that of the puddle rolls above de- 
 scribed. Taken strictly, one would expect a less draw for 
 small bar-iron, because the latter is pretty solid, and espe- 
 cially because it is of great importance that the finished bars 
 should be free from imperfections ; but the Styrian iron bears 
 quite a heavy draw, in spite of its frequent steel-like quality, 
 and the absolute draw of the puddle rolls above described 
 was not very great. At other mills, the draw of the same 
 rolls is frequently much lighter ; in which case single passes 
 can be often skipped, where experience has shown this to be 
 admissible. 
 
 If the grooves are large and deep they weaken the roll 
 very much, by reducing its sectional area, and this is espe- 
 cially the case when they are near the middle, as this is the 
 weakest point. In such cases the diameter of the roU 
 must be so great that its diameter in these grooves is greater 
 t'lan the diameter of the neck, and the passes themselves 
 should always be placed next the neck. For these reasons 
 the largest Gothic passes are often placed at one end of the 
 rolls, then the second, and so on ; but in the case of small 
 Tolls, or any in which the diameter may be proportionately 
 too small, it is advisable to place the first and second passes 
 at one end, and the third at the other, the succeeding passes 
 decreasing in size toward the first two. Many roUers think 
 that the first or welding passes should be nearer the centre 
 of the roUs, as the slag which is squeezed out of the pile 
 
27 
 
 might get into the journals ; but it is much more important 
 to avoid breakages by making the rolls strong, than to avoid 
 a hypothetical injury to the necks or journals, especially as 
 the latter can be easily prevented by placing an iron shield 
 or cinder-plate between the rolls and journals. Sometimes the 
 roll is especially notched to receive the cinder-plate, while it 
 is often the case that the end-fillet of large rolls sufficiently 
 protects the neck. 
 
 The use of Gothic passes in three high trains for small bar 
 is quite frequent, and very advisable, as the arrangement of 
 the passes presents no difficulty. The construction of the 
 passes remains the same, but it is necessary to make the 
 diameter of the top roll about of an inch greater, and that 
 of the bottom roll about xV of an inch smaller than that of 
 the middle roll. In this case the guards rest on bearing-bars 
 before the middle and lower rolls. If hanging guards are 
 used, the top roll should also be about | inch smaller than the 
 middle roll. In a forge near Leo ben the three high system 
 is thus used for the puddle rolls, and with marked success. 
 
 The last Gothic pass represented in Fig. 7, Plate I., has a 
 diameter of nearly | of an inch (or ^ inch length of side) ; 
 but it is seldom that the diameter is less than one inch, as 
 the pass, when so small, would not roll accurately enough, 
 especially if the sides were much curved. When the drawing 
 passes must be so smaU, it is better to use oval and diamond 
 passes alternately. 
 
 § 12. Box passes are used as welding passes, when the pile 
 is large, or its form varies considerably from a square. Such 
 passes are, of course, always open ones, with rounded angle?, 
 while the sides of the body-fillets form an obtuse angle with 
 the bottom of the pass, as shown in Plate II., Fig. 16. The 
 grooves of the top roll are notched or furrowed, in order that 
 they may take a better hold. 
 
 When the section of an open box pass does not dififer ma- 
 terially from a square, it is often so constructed that the pile 
 may be passed once, turned quarter over, and returned 
 through the same pass, which has been somewhat closed. 
 In this way a few passes may be saved, and sometimes the 
 
28 
 
 pile is rolled thrice through the same pass, which is opened 
 wide at first, and gradually closed ; but the economy in rolls 
 and turning by no means compensates for the increased 
 labor and longer time required for rolling. Such passes are 
 shown in full size by Fig. 19, on Plate 11., as used in a 
 Carinthian mill. But, as formerly observed, such a construc- 
 tion should be resorted to only in cases of necessity, as the 
 proper relation of draw to side pressure cannot in such cases 
 be at all regarded. 
 
 Box passes cannot strictly be considered as drawing 
 passes, although their draw is often very heavy, because, 
 in those cases where they are used alone, the pile is shaped 
 according to the form of the finished product, so that those 
 passes which form the first two or three are in reality shap- 
 ing passes. For shaping as well as for finishing, it is best 
 to construct the flat pass as closed, and to round off the 
 angles, not only to avoid unequal cooling of the metal, but 
 also to prevent the former of a subsequent pass from 
 forming a fin on the corners of the bar. In order to obtain 
 a finished bar with sharp, square corners, it is not only 
 necessary to make the last pass rectangular, but also to 
 make the angles of the pass next before the last so far 
 square, as determined by experiments, that the bar, turned 
 half over (180 degrees), will have sharp corners. When the 
 bars are to be piled and re-rolled into smaller sections, the 
 angles of the last pass should be left rounded, as rounded 
 corners bear heating better than square ones. For this 
 reason, puddle or mill-bar grooves are generally turned with 
 rounded angles. 
 
 It is important, in the successive use of flat passes, not 
 only to have a proper draw, but to determine whether the 
 pass shall widen or not as its depth decreases, and if the 
 former, how much. As the bar cools between the passes, 
 it must also contract, and becomes, therefore, a little nar- 
 rower at every pass, even though the amount of shrinkage 
 is barely appreciable.* 
 
 * The shrinkage from the rolling heat to an entire coldness is generally 
 taken as ; it varies somewhat with the kind of iron, sometimes 
 
29 
 
 When the flat bars are quite wide, this dimimition of the 
 width as the heat falls is very plain, and the tar is easily- 
 rolled through passes of the same width. When, therefore, 
 flat bars must be made, from 10 to 12 inches wide, for 
 very large piles, it is often well to use a single closed flat 
 pass, through which the bar is passed several times, 
 while the top roU is screwed down after each passage. 
 When, however, the passes have the same width, the sides of 
 the grooves must flare out from the bottom a full 
 of an inch per inch of height, in order to render the passage 
 of the bar easy, while the former is made rectangular, and 
 of such width that the top roll can be screwed down a 
 little. 
 
 Although the use of flat passes of the same width reduces 
 the number of passes and rolls, yet, as above observed, they 
 are seldom used, because the rolling is rendered more diffi- 
 cult on account of the increased difficulty in introducing the 
 bar into the pass ; and also, because it is almost impossible 
 to avoid fins. 
 
 The draw of successive flat passes is generally in the case 
 of welding and drawing, or, as it were, shaping passes of large 
 size, in the proportion of 5 : 4, or, if the quahty of the iron is 
 good, as 4 : 3 or 3 : 2. The draw of smaller passes is quite 
 different, and may become, at the last pass, for a bar which is 
 only a few hues thick, as great as 2 : 1, or even heavier ; but 
 the absolute reduction of height amounts in this case to a 
 few lines, while it may often be 1 to 1| inches in the larger 
 passes. It follows, therefore, that the draw or lengthenmg 
 out of the bar increases greatly as the size of the bar 
 diminishes. 
 
 as much as to being estimated, as in the case of irons made from 
 clay ironstone, therefore the more impure kinds. All finishing passes 
 must be made greater by this amount than the finished section requires. The 
 distance between the saws must be, in the same proportion, greater than the 
 desired length when cold. The influence of the varying temperature at 
 which the rails are sawed upon the final length is very marked ; differences of 
 haK an inch and more are frequent when the temperature at the saws is not 
 kept pretty nearly the same. 
 
30 
 
 The increase of width amounts in large passes to about 
 to I of an inch, and in the smaller passes to to of an 
 inch, it being taken for granted that the shape does not 
 require any greater widening than this. Heavy and wide 
 flat bars of 7 to 9 inches width are rolled from a pile, and in 
 order to insure sound welds, one or two edge passes are used, 
 which must naturally be made disproportionately wider, viz., 
 I to I of an inch, because these passes contract toward the 
 bottom. Three high rolls are often used for rolling small 
 and narrow flat bars, but on account of difficulty in raising 
 or lowering them, etc., they are seldom applied to rolling 
 bars more than 3 inches wide. Thin hoop-iron, after being 
 finished, is passed through poHshing rolls, in order to obtain 
 a handsome surface ; these rolls should, if possible, revolve 
 only about half as fast as the finishing rolls. 
 
 The body-fillets of rolls with flat passes are not, like those of 
 the Gothic form, much stronger {i. e. thicker) at the bottom of 
 the pass ; they must, therefore, be made from 2 to 4 times as 
 wide as those usual with the Gothic form. For this reason, as 
 well as on account of their considerable width, these passes re- 
 quire many rolls. Now, as flat iron is desired of every possi- 
 ble width, a mill in which it was expected to roll every width 
 would be obliged to keep on hand a gi-eat many rolls. The 
 necessity of this great stock may be partly obviated by the 
 use of step rolls ; but the best means of avoiding it is the use 
 of the Universal Mill arrangement, with two horizontal and 
 two smaller and vertical rolls, by means of which flat iron of 
 almost every size, and especially of the larger sizes, can be 
 easily rolled.* 
 
 § 13. Figure 8 on Plate I. represents a pair of rolls with 
 flat passes for puddle bar, 4, 3, and 2 inches wide and 1 
 inch thick ; these rolls are in use in a Silesian mill. The 
 proper roughing rolls for this finishing pair are the puddle 
 
 * The invention of the Universal Mill is conceded in Europe to Herr 
 Daelen, the Engineer of the Ironworks at Horde in "Westphalia, and was first 
 publicly described in 1856, by Tunner, in the annual volume of Essays pub- 
 lished by the Austrian Mining Schools. 
 
31 
 
 roughing rolls, witli Gothic passes, whicli were described in 
 § 11 and drawn in Figure 3, Plate I. The Gothic pass last 
 used must of course be a little narrower than the first flat 
 pass, in order to insure an easier introduction of the bar ; 
 with the same object the latter is also sometimes rolled a 
 second time through the last Gothic pass, after being turned 
 quarter over, in order that both diagonals may be of the 
 same length. Figs. 9, 10, and 11, on Plate I., show in full 
 size the passes of the rolls in Fig. 8, the pitch line being 
 also shown. As the bars are intended for subsequent piling, 
 the angles of all the grooves in the lower roll are rounded off. 
 
 Fig. 12, on Plate I., shows, in full size, passes which are 
 used in a Styrian mill to roll flat bars (mill bars), 7 inches in 
 width ; the pitch line is also shown. Five passes are used, the 
 second of which is an edge pass. The forge hammer pre- 
 pares the ball for the first pass, which is 7| inches wide and 
 2| inches deep. The reduction of the bloom in the first pass 
 varies somewhat, as the hammer work is not very accurate. 
 The edge and also the third pass have a draw of about | of 
 an inch, and very slight increase of width. The fourth is 
 about an | of an inch wider than the third, and has a draw 
 of 1% of an inch ; a heavier draw would not be advisable, as 
 the bar has by this time cooled considerably. If puddle bars 
 are desired much wider than the above, it is necessary to 
 draw them out under the forge hammer. 
 
 In Fig. 12, on Plate I., the passes of a pair of rolls are re- 
 presented without their rolls, and as this will be often 
 necessary, it will be useful to describe the method of laying 
 out rolls corresponding to any given set of passes. These 
 passes (viz., those of Fig. 12) are quite wide, and the draw 
 is heavy ; therefore, as the rolls inust have considerable 
 strength, the necks must be strong — say 9 inches in diameter 
 — as drawn in similar cases in Fig. 8, Plate I., and Figs. 16 
 and 17, on Plate II. The deepest groove is the second, or 
 edge pass ; therefore the latter must be placed next the neck, 
 while the diameter of the roll at the bottom of the groove 
 must be greater than that of the neck itself (§ 11). The 
 height of the pass above the pitch line is, as shown in Fig. 
 
32 
 
 12, Plate I., a little over 3 inches ; and since this pass, as well 
 as the other four, are best constructed as closed passes, the 
 collar of the under roll must project about 3^ inches above 
 the pitch Hne, and the bottom of the groove in the top roll, 
 which receives these collars, must be at least 3| inches above 
 the same line, in order to allow the necessary play. As 9 
 inches was adopted as a suitable diameter for the necks, 
 the diameter for the roU, at the bottom of the edge pass, 
 must be 10 inches, or, in other words, have a radius of 5 
 inches. The centre line of the top roll must, therefore, be 
 placed 5-f-3|=8^ inches above the pitch line. The edge 
 pass extends 3 inches and 8| twelfths below the pitch line 
 (as in Fig. 12), and the diameter of the bottom roll at the 
 bottom of the groove, must be at least 10 inches (5 inches 
 radius) ; therefore the centre line of the bottom roll must lie 
 8 inches and 8| tweKths below the pitch line. The centre 
 line of the top roll should, therefore, be placed 2| twelfths of 
 an inch higher (in order that the distance of both centre 
 lines from the pitch line may be equal), otherwise the radius 
 of the bottom roll, at the bottom of the groove, would have 
 to be made 2^ twelfths smaller, if for any reason it were not 
 advisable to raise the edge pass 1 1 twelfths nearer the centre 
 line of the top roll, as is sometimes done. Farther, the col- 
 lars of the bottom roll rise on each side of the edge pass, 3| 
 inches above the pitch line; therefore the radius of the body of 
 the bottom roll must be 8 inches and 8^ twelfths and 3| 
 inches=12 in. and 4^ of an inch. The former of greatest 
 diameter on the top roU is that which closes the fifth pass 
 (Fig.|12) ; it wants i^ch of reaching the pitch line, there- 
 the body of the top roll must be made with a radius of 8 inches 
 and ^ twelfths (8 in. 8^ twelfths— t\ in. =8 in. 3| twelfths). 
 These remarks apply only to those rolls which are cast cylin- 
 drically, without the grooves in the rough. Where, however, 
 the grooves are roughly cast on the roll, the collars are made 
 of varying diameter, as required by the construction. Con- 
 cerning the necessary length of the body of the roll, it must 
 be remembered that the interior collars must be at least four 
 times as strong as the body-fillets of Gothic passes, which are 
 
33 
 
 ^ in. wide at the top ; therefore, the former should be, in this 
 case (as they are very high), at least 2 in., or better ^ in. 
 wide, while the end collars should be 3.| in. wide. The col- 
 lars, then, require a length of 21X^+3^X2=17 mches, and 
 the passes take up (Fig. 12), 7"6"'+2"7^"'+6"10i"'+7"+7" 1-^ 
 =31"! The bodies of the rolls must, therefore be made 
 17"+31"=48" long, neglecting the |". All the data necessary 
 for constructing proper rolls for the passes of Fig. 12 are 
 thus fixed, and it is only necessary to transfer them to 
 paper. 
 
 Fig. 15, on Plate II., represents a pair of puddle rolls in 
 use at a "Westphalian mill; they contain the necessary 
 roughing and finishing passes for a puddle bar 3| inches 
 wide and | inch thick. These rolls are short, being only 49 
 inches between the necks. But three Gothic roughing passes, 
 as shown in the figure, are too few, unless the puddle balls 
 vary little in size, and are very carefully bloomed under the 
 hammer. It is usual to employ five roughing passes, in 
 which case the body of the rolls must be over five feet long, 
 which length necessitates very strong rolls, especially if the 
 grooves are deep as in these roUs, since any great length 
 of weak body causes frequent breakages. If additional 
 passes for bars of different widths were to be turned on 
 these rolls, the length would be too great for safety. Such 
 puddle rolls have the advantage that a single pair of rolls 
 and housings contain all necessary passes, thus shortening 
 the train ; they are, accordingly, often used in large mills 
 where puddle bar of one width is rolled the whole year 
 round. If, however, the size of puddle bar to be rolled 
 varied frequently, it would be troublesome to change rolls 
 of such great weight and length, and too large a stock of 
 rolls would be necessary. In such cases, and, indeed, gen- 
 erally, it is preferable to divide the roughing and finishing 
 passes upon two pairs of rolls, as the roughing rolls need 
 not then be disturbed (the number of passes being sufiicient 
 for several sets of finishing passes), while a single pair of 
 finishing rolls of moderate length, as in Fig. 8, Plate L, 
 
34 
 
 may contain passes for three widths of puddle bar, which is 
 usually sufficient for current manufacture ; and if the puddle 
 finishing rolls must be changed, such rolls as these can be 
 far more easily and quickly changed than those in which 
 roughing and finishing passes are on the same roll. In 
 some mills a second pair of finishing rolls is used to avoid 
 frequent changes, but one set is hkely to be so seldom 
 used that this practice cannot be recommended, as the 
 power consumed, even in running a pair of rolls hght, is 
 considerable. 
 
 An example of the use of flat passes in roUing wide flat 
 iron, is presented on Plate II., in Figs. 16 and 17, which are 
 respectively the roughing and finishing rolls of a set used in 
 a Silesian mill for rolling bars 7^% inches wide. The dimen- 
 sions printed on the drawings render all description unneces- 
 sary, further than a reference to § 12. On the top rolls, not 
 only all the roughing, but also the first two finishing passes 
 are roughened or furrowed, that they may take a firmer hold, 
 and the two latter passes both at right angles to and parallel 
 with the centre line of the rolls. When passes not roughened 
 as above will not take hold, the bar must be forced forward 
 by strong blows with the buggy at the same time that sand 
 is thrown on its surface. The thickness of the flat bar can 
 be varied by raising or lowering the top roU within a certain 
 limit, which is, in the present instance, from to 1% of an 
 inch. 
 
 Figure 18, on Plate II., represents a pair of rolls with flat 
 passes, for bars 2*- to 3 inches wide, as used in a Silesian 
 mill. The roughing rolls corresponding to these would be 
 those drawn in Fig. 5, Plate I., since, as formerly re- 
 marked, the proper pass is used as the last, and the bar 
 is passed twice through it before entering the first finishing 
 pass. 
 
 As an example of the three high system for flats, a set of 
 finishing rolls for bars, 2 inches and 3| twelfths and 2 y\ 
 inches wide, and with five grooves respectively, are drawn in 
 Fig. 27, Plate II., while their passes are represented sep- 
 
35 
 
 arately and in full size in Figs. 28 and 29. Fig. 20 further 
 represents the three finishing rolls for bars \l, -Vsr? -Vitj tI > 
 •fl, and of an inch wide, with three passes for each size ; 
 the passes being shown, in fuU size, in Figs. 21 to 26. Both 
 sets of rolls are iV of fiiH size. The draw is, in these in- 
 stances, heavy. The rolls described in § 11 would serve 
 the above as roughing rolls, and would be constructed 
 similarly as three high sets. 
 
 The step-rolls drawn in Fig. 33, Plate III., are used in 
 the manufacture of flat bars, and they will be here described 
 (since they are, as it were, a variety of flat pass as far as 
 their use is concerned), together with the guide arrangement 
 represented in Fig. 34. So far as the rolls themselves are 
 concerned, which iu position and use are similar to grooved 
 rolls, the drawing is sufficient without further explanation ; 
 the guide arrangement, however, requires the following de- 
 scription : 
 
 The bar a, which lies between the two housings, horizontally, 
 and at the proper height, is the bearing bar, to which the 
 guide arrangement is fastened by the clamp-screws b b. In 
 the rectangular frame or box of the arrangement, two side 
 plates, c c (front view), are placed, each of which is provided 
 on one side with 4 set screws. By means of the middle 
 screw d, each of the plates can be moved outwards (i. e. 
 away from the other), and the space between the two can 
 thus be exactly fitted to the thickness of the bar which is in- 
 troduced on edge ; the three screws / hold each plate firmly 
 in the required position. In order that the bar may be 
 guided as exactly as possible, and thus prevented from bend- 
 ing over sidewise, the ends of the plates are cut to the shape 
 of roUs in order to reach weU in between them and hold the 
 bar as long as possible. This guide is usually placed before 
 one of the divisions of the step roUs, or it may be also set up 
 before a pair of plate or polishing roUs, the position of which 
 to each other can be altered at will, and which are usually 
 set in the same Une as the step-rolls, and alongside the same. 
 The bar must usually pass twice through this adjusting 
 
36 
 
 arraiigement, and, as a rule, immediately before beiag rolled 
 through the last finishing pass. Step rolls are most useful 
 in finishing the edges of hoop iron of various widths, as well 
 as the edges of all thin bars, on the corners of which fins 
 are likely to be formed. These rolls are always chilled for 
 polishing, and obviate, by their form, the otherwise frequent 
 raising and lowering of the top roll of a pair, while thin flats 
 of different widths can be readily finished in them at the same 
 time. 
 
 § 14. Diamond passes, especially those of large size, are 
 used as finishing passes, while Gothic passes are used as the 
 corresponding roughing passes. In the case of small, square 
 iron, however, the ordinary Gothic form is not accurate 
 enough, and the more inaccurate the greater the curve of its 
 sides ; hence, either a Gothic pass, with sides very slightly 
 bent (Fig. 7, Plate I.), may be used as roughing passes, or a 
 pass, the sides of which are quite straight, as ia Fig. 14, 
 Plate I., which last form is classified with the diamond 
 passes, although it differs quite widely from the true diamond 
 form. For rolling very small square rod, and round or wire 
 rod, the first roughing passes are usually Gothic ; the fol- 
 lowing, however, are diamond and oval alternately; the 
 latter are interpolated as drawing passes. 
 
 Diamond passes are, with few exceptions, constructed as 
 open passes. A bar of square section might be regarded as 
 a thick and narrow flat bar, and rolled accordingly. This is, 
 indeed, sometimes done in practice, but merely exceptionally, 
 and the bar must be passed twice or oftener through polishing 
 roUs, opened to the thickness of the bar, and must be turned 
 90 degrees (or quarter over) each time. But as a rule this 
 practice would not be advantageous, and need not be further 
 considered. 
 
 The diamond passes are always so constructed and turned 
 that one diagonal stands at right angles to the centre line of 
 the roll, while the other hes between the rolls. The bar is 
 rarely finished, or more properly polished, in a single pass, 
 
37 
 
 but, according to circumstances, in 2 to 4 passes ; and, iiie^e- .-' 
 fore, in order to avoid fins on the bar, and to improve tlie 
 bold of the passes, the vertical diagonal is made somewhat 
 shorter than the horizontal. 
 The vertical diagonal is 
 shortened, or, what is the same 
 thing, the horizontal diagonal 
 is lengthened by turning the 
 angle a in the adjacent Fig. 
 11a, not as a right angle, as an 
 exactly square section would 
 require, but as an angle of 
 92|- degrees- 
 
 This obtuse angle of 92| degrees is used with an addi- 
 tional purpose, viz., to prevent the unequal contraction of a 
 bar of square section and the sharp comers caused thereby. 
 For if the bar came out of the last pass with rectangular 
 corners, these would become sharp and acute angled on 
 cooling ; because the corners which have 
 been already disproportionately cooled in fig. 11b. 
 rolhng become cold sooner, and conse- 
 quently contract less, than the middle 
 part, which has been the hottest part, and 
 remains so longest on cooling. In this 
 case the side of the cold bar would be 
 concave, and the angles of its corners 
 more or less acute, as shown in the adja- 
 cent Fig. 11b. 
 
 The nearly finished bar is usually roUed twice through the 
 last pass, being turned quarter over each time. In this way 
 the variation of the section of a bar from an exact square is 
 rendered very inconsiderable, even in large bars, since both 
 diagonals thus become equally long. In the case of very 
 small square rod a second rolhng through the pass would 
 occasion too long a detention, and as the pass must be pro- 
 vided with very accurate guides, it is better to adopt the 
 following method of securing a correct section. This method 
 is the same as that followed in rolling wire rod. The pass 
 
38 
 
 Fig. 13. 
 
 next before tlie last is turned with an angle of 110 de- 
 FiG. 12. grees, as is also customary 
 
 in most diamond drawing 
 passes. The form of the 
 pass is so regulated with 
 reference to that of the 
 preceding and following, 
 that the body-fillets do not 
 touch, as is shown in Fig. 
 12. A few trials will show 
 the exact position of the rolls at which the bar wiU exactly 
 fill the last pass, and leave no fin after it has gone through 
 the above pass (next to last) and has been turned half 
 over. 
 
 In order to construct the above finishing passes with angles 
 of 92|", or the drawing passes with those of 110°, without undue 
 
 waste of time, it is well to draw 
 the angle of 92|o or 110°, as 
 the case may be, once on paper, 
 as in adjacent Fig. 13 ; bisect 
 the angle with the line a a, and 
 from the points of the angle a 
 lay off upon this line, and in the 
 proper relative proportion to the 
 the others, the half of the vertical diagonal of each succes- 
 sive pass, as a a, a &, a c, etc. ; through this line, and 
 at right angles to it, draw the lines o c, n h, a m, etc., 
 which are the halves of the horizontal diagonals, and are of 
 the required length. The individual passes are most con- 
 veniently constructed by means of these diagonals. In 
 tumiQg these passes it is merely necessary to grind the point 
 of the tool exactly to an angle of 92^° (or 100°, as the case 
 may be), so that the proper proportion of the diagonals to 
 each other may be preserved. 
 
 When the larger diamond passes are used only as finish- 
 ing passes, the small difference in the length of the two 
 diagonals makes it necessary that the draw be very light. 
 Keckoniug by the length of the side of the section, a reduc- 
 
39 
 
 tion of this side by {{. e. reduction of area as 100 : 81), 
 supposes quite a heavy draw, as the reduction, as measured 
 on a side, is often and even less. The fact that bars of 
 very shghtly different sectional areas must be rolled in the 
 finishing roUs, makes such a slight draw especially useful ; 
 the draw may with advantage be so light, that individual 
 passes may be readily skipped. Small changes of area may 
 of course be made by raising or lowering the top roll. 
 
 As the difference between the vertical and horizontal 
 diagonals of diamond drawing passes is considerable, they 
 may be made with so heavy a draw that they will just take 
 the bar. This great difference between the diagonals also 
 permits considerable play between the rolls, viz. : to of 
 an inch, as in Fig. 14, on Plate I. ; this play allows the draw 
 to be somewhat changed, as the behavior of the bar may 
 require, while the readiness with which the change of draw is 
 made makes it easy to skip one or more passes if desired. 
 The extent to which such skipping may be practised 
 depends on the quality of the iron. This purely empirical 
 method leads in general to pretty nearly the same draw, as 
 will be shown when oval and round passes are considered. 
 This draw is shown in Fig. 35, Plate III., as applied to roll- 
 ing wire rod. 
 
 Three high rolls are frequently used for small, square iron, 
 when it is not desired to adopt the " high speed " plan, by 
 which a greater number of housings and rolls are rendered 
 necessary, in order to keep the bar always in two or three 
 passes at the same time. This system has already been 
 mentioned in § 6, and will be specially described in § 17. 
 
 § 15. It will be sufficient to give merely two examples of 
 this kind of pass, viz., Fig. 30, Plate II., which represents a 
 pair of rolls for large square bar, in ^ full size ; and Fig. 32, 
 which is a single roll for small square rod, the latter being to 
 ^ fuU size on account of the small grooves. Instead of show- 
 ing the passes of the former pair of rolls in full size, a 
 graphical representation of the ratio between the horizontal 
 and vertical diagonals is annexed in Fig. 31. The above 
 
 4 
 
40 
 
 rolls are, of course, finishing rolls ; therefore the angle of the 
 latter diagram is 92 1 degrees. These should be, like polish- 
 ing rolls, always chilled rolls. 
 
 One detail, which has not yet been noticed, should be 
 remarked in Fig. 30, Plate 11., viz., a construction which is 
 usual for large rolls with open passes, and was formerly 
 especially frequent, when it was necessary, for the sake of 
 accurate work, to prevent the rolls from being forced endwise, 
 and to do this by means of the rolls themselves. To this end 
 the collars at the end of the bottom roU are made of greater 
 diameter than that part of the top roll opposite them, and 
 the former revolves in the half open depressions in the top 
 roll, as shown at each end of the body of the roll at a and h. 
 This construction is, however, seldom applied to small rolls 
 of this kind, because these require to be placed with great 
 accuracy, and special set screws in the chocks allow the most 
 reliable and accurate adjustment, and are now generally used, 
 even for the largest rolls. 
 
 Fig. 13, Plate I., represents, in -j-V fuU size, a pair of rough- 
 ing rolls, in use at a Westphalian mill, while the correspond- 
 ing passes are shown in full size in Fig. 14. The angle of 
 the roughing, or rather drawing passes, is 110 degrees. All 
 necessary explanations will be found in § 14. 
 
 § 16. Oval passes are chiefly used as drawing passes, or as 
 shaping passes for small bar, or especially for drawing wire 
 rod, as thek form, as stated in § 7, is well adapted to the 
 reception of a very heavy draw. In the first of these in- 
 stances, oval alternate with diamond passes, and accurate 
 guides must be used before each pass. The draw, in this 
 case, varies from i to | the height of the bar which is about 
 to be rolled. These oval passes are always open, and the 
 rolls, which are generally 8 to 10 inches in diameter, are so 
 turned, after the passes are finished, that a play of 4V to 4^ 
 of an inch is left for adjusting the rolls in the housings. 
 
 In constructing these passes a circle is used as the ground 
 form. To arrive at an oval, however, there are many differ- 
 ent ways, which vary according to the size of the passes. 
 
41 
 
 
 
 
 3f- 
 
 e 
 
 
 -A — ^ — 
 
 — ^— k. 
 
 
 
 
 f 
 
 
 
 d 
 
 
 The following is a method used by Talbot as the best con- 
 struction for medium round iron, ^ in. to 1 inch in diameter. 
 
 The circular section of the required bar is first drawn, as 
 in Fig. 14 A, which circle represents the last pass, to precede 
 which an oval pass is to be constructed. 
 
 Divide the diame- 
 ter of the circle J ^'S- 14. 
 into 3 equal parts ^ 
 and with two of 
 them as radius de- 
 scribe a second cir- 
 cle B, the diameter 
 of which is c d. Di- 
 vide this diameter 
 also at e and / into 
 3 equal parts, and through e and / draw the lines g h and 
 i k at right angles to the diameter c d. The arcs g oh and 
 idk form when placed together the desired oval pass. Fig. 
 15 represents the oval drawn over the final 
 round pass, in order to show the relation of 
 the draw to the corresponding "spread." 
 This relation is then finally adjusted by trial 
 in the housings, as is more specially discussed 
 in § 17; the oval and round pass are of course 
 on different pairs of rolls and must be fur- 
 nished with accurate guides. 
 
 The following construction, which was in- 
 vented by Talbot, is well adapted for the larger sizes of round 
 iron ; it is shown in Fig. 16, and is as follows : Describe a 
 circle of a diameter (a h) equal 
 to the height of the pass ; in 
 this draw a horizontal and 
 vertical diameter. The radius 
 of the circle is divided into 5 
 equal parts, and the horizon- 
 tal diameter is extended on 
 each side by the length of one 
 of these parts, to the points 
 
 Fig. 15. 
 
 Fig. 16. 
 
42 
 
 0 and /; also a space equal to one such part is laid off on the 
 vertical diameter from the centre to the points d and c, and 
 the half of this space marked in the points x and y. Through 
 the points c and d draw the lines i k and 1 1 parallel to o /• 
 From 7/ as a centre, with the distance y a as radius, describe 
 the arc nam, and similarly from x, the arc g b r. The arcs 
 n 0, 0 q, rf, and / m are then described with the radius of 
 the large circle. The pass is then complete, as the figure 
 shows. 
 
 As, however, passes of a form similar to that of Fig. 16 
 are almost round, and are used to finish round iron, they 
 might almost be counted as being round passes, though 
 properly belonging to the oval class. Two such passes are 
 commonly used successively in finishing heavy round bars, 
 which are rolled twice through each pass, or, better, once 
 through the first and three times through the last, being 
 turned each time one-quarter over.. The draw of these passes 
 is therefore very light, amounting to about diameter, 
 or the height of the pass. 
 
 An example of the use and form of very flat oval passes 
 is given in Fig. 35, Plate III., and described ia the following 
 article. 
 
 § 17. Fig. 36, on Plate III., represents a pair of finishing 
 rolls used in a Silesian mill ; they contain passes for round 
 bar of from 1.63 to 0.8 of an inch in diameter (in addition to 
 octagon passes), which are in reality oval, and constructed 
 according to the rules laid down in § 16, although they seem 
 round on account of the small scale of the drawing. The 
 purpose with which they are presented is merely to give a 
 practical example, in which the size and construction of such 
 rolls, as well as the draw of the passes, may be plainly seen ; 
 the accurate construction of the passes is that given in § 16. 
 In the drawing, the fillets appear rectangular, and seem to 
 touch the corresponding parts of the other rolls, while in 
 reality they are slightly rounded, and have ^ to ^ig- of an inch 
 play, to prevent the corners from breaking off (§ 10) ; the 
 scale is, however, so smaU that these details do not admit of 
 being shown. 
 
43 
 
 Very flat oyal passes are especially adapted to rolling wire 
 rod. They tlien alternate with diamond passes, while a true 
 round pass is used to finish. Fig. 35, Plate III., shows these 
 passes in fuU size and natural succession, as they are used in 
 a Carinthian mill in rolling wire rod -^^ in. thick, out of pud- 
 dle bar of medium quality. The following explanation is 
 necessary to a correct understanding of the drawing : 
 
 The l|-inch bar is first rolled at a welding heat through a 
 set of three high rolls with Gothic and diamond passes, going 
 altogether through 8 to 9 passes ; then, before it has entirely 
 come through the last pass of the first set of rolls, it is intro- 
 duced into the first pass, marked 1, of the second set of two 
 high rolls with diamond drawing passes. The bar passes 
 thence into the oval pass 2, which is turned, together with 
 several others of the same kind, upon a set of two high rolls 
 in the third housings ; these latter rolls are driven in a con- 
 trary direction to that of the second pair. Thence the bar 
 is taken back into the diamond pass 8 in the second pair, 
 thence into the oval pass 4 in the third pair, again back into 
 pass 5 in the second pair, then into pass 6 in the third pair, 
 and back into pass 7 in the second pair. By this time the 
 bar has become so long that it is being rolled into two or 
 even three passes at the same time. The last oval pass 8, as 
 well as the finishing round pass 9, are mounted in separate 
 housings, iu order to ascertain, by special trials, the exact 
 position of the rolls at which the sectional area of the oval 
 bar is exactly large enough to fill out the round pass com- 
 pletely, without forming fins, and so that the section of the 
 finished rod does not vary perceptibly from a perfect round 
 as it comes out of the round pass 9. The finished rod is 120 
 to 150 feet long. The number of passes through which the 
 rod is rolled is 8 (or 9)-f 4+3+1+1 = 17 or 18, while this 
 " high speed" method of rolling reduces the time necessary 
 for the whole operation to not much over one minute. The 
 roUs have a diameter of 9 inches, with the exception of those 
 containing passes 8 and 9, the diameter of these latter rolls 
 being about 1\ inches. They are, however, only 7 inches 
 long between the necks, while the bodies of the former are 
 
36 to 40 inches long. All the rolls make about 250 revolu- 
 tions a minute. The speed of rolls for rolling wire rod 4 in. 
 thick has been recently much increased. These rolls are 6 
 to 7^ inches in diameter, and make 400 to 500 revolutions in 
 the minute. This necessitates, however, very expert rollers 
 and flatter passes, with a heavier draw. 
 
 It might seem that the diamond passes in Fig. 35 should 
 have been turned with an angle of 110 degrees, since they 
 are drawing passes, but as they are used in conjunction with 
 oval passes, in which the bar is made very flat, it suffices to 
 make the angle 90 or at most 92 J degrees, as above. 
 
 § 18. The construction of round passes has been partially 
 discussed in the course of the description of various oval 
 passes in § 16 and § 17, because the heavier kinds of round 
 iron are rolled exclusively in oval passes, or because, as 
 in the case of very small round rod, an oval pass is used 
 next the last, which stands in close relation to the round 
 pass, which finishes. In some mills, however, it is customary 
 to use true round passes in the finishing rolls for medium 
 round bar of greater diameter than | of an inch ; these 
 passes vary somewhat from the form described in § 17. 
 
 A construction of this kind, used by Talbot, is represented 
 in Fig. 17. The diameter (a b) of the circle which repre- 
 
 Fig, 17. 
 
 / \ 
 
 sents the area of the pass is divided into 6 and also into 8 parts 
 or, rather, each half or radius respectively into 8 and 4 equal 
 parts. On the continuation of the diameter a h, the distances 
 
46 
 
 hdandac are laid off equal to | of tlie diameter, and from a 
 and 6, arcs intersecting the circle a.t eg,f h, are described, with a 
 radius equal to 4 of the diameter. The arcs e c, c f, h d, and 
 g d, are described with a radius equal to | the diameter of 
 the circle, and the pass is thus complete. 
 
 Another and older method of drawing such passes is shown 
 in Figure 18, as described long ago by Karsten. It consists 
 in describing, from the ends of the 
 horizontal diameter a b, with the 
 radius h c, the arcs cpnd and comd. 
 Each of these arcs is then divided 
 into six equal parts, and the points 
 e, /, g and h are laid off from a h 
 upon the circumference of the cir- 
 cle, at a distance equal to one of 
 these parts. The arcs e'k,h f, etc., 
 are described respectively from m o, 
 etc., with the respective radii equal to me; these arcs com- 
 plete the pass. The arcs comd and cpnd may be divided 
 into 6 equal parts by describing, with the radius b c, from the 
 points c and d, similar arcs a p ob and anmb (not shown) ; 
 these arcs intersect the first in o, p, n, m, and the distance 
 from any of these points to the line a b will be ^ of the above 
 arcs. 
 
 Constructed in this way, the greater part of the pass is a 
 true circle, which is slightly enlarged at the horizontal diam- 
 eter in order to avoid fins. The iron forced into the resulting 
 depressions in the pass is rolled into the bar again at the 
 next pass, thereby rendering it easy to introduce the bar into 
 the pass, and improving the nip of the latter. 
 
 After grooves formed in the above way have been accu- 
 rately turned upon the rolls, it is necessary to turn off the 
 body-fillets afterwards till a play of from to -i-^ of an inch 
 is obtained between the rolls, in order to be able to adjust 
 the passes properly. When rolls are turned with grooves of 
 the form described in § IT, and shown in Figure 35, Plate III., 
 the body-fillets are afterwards turned off so far that a play 
 of to ^ of an inch is left between the roUs, in order to be 
 
46 
 
 able, by the adjustment of the rolls, to bring tbe passes into 
 tlie desired truly round form. 
 
 The constructions described in Figs. 17 and 18 are used 
 without guides ; for if a guide is used, the bar must be fin- 
 ished at a single passage, as the guides do not possess the 
 necessary accuracy when the bar is very nearly round. The 
 distinction between guide rolls and those without guides is 
 not a sharp one. In many mills, all roimd bars under an inch 
 thick are rolled with guides ; in others, rounds as low as f of 
 an inch, or even less, are rolled without guides, the bar being 
 guided by hand. The diameter and the required length of 
 the bar determine the kind of rolls to be used ; for the longer 
 the bar becomes, the more uncertain becomes its guidance 
 by hand, even though a second roller helps to guide. 
 
 Eoughing rolls for large rounds are similar to those shown 
 in Figs. 5, 6, 7, 13, and 14, on Plate I., all which have been 
 already described. Finishing rolls for large rounds have 
 been described in the first part of § 17, and represented in 
 Fig. 36, Plate HI. 
 
 § 19. Polygon passes are employed almost exclusively in 
 the hexagon or octagon form, to finish bars of similar sec- 
 tion. The corresponding roughing rolls are, like those for 
 rounds, furnished with Gothic passes. When the section is 
 large, as for instance 1^ inches between the sides, and when 
 the comers are to be rolled sharp and the sides smooth, it is 
 best to use three shaping passes, a plan which is preferred by 
 Mr. Baildon, of the Donawitz mill, near Leoben. 
 
 Fig. 19. 
 
 ■<©> "0" 
 
 ' These are shown in Fig. 19, in which the height of one 
 pass is always somewhat less than the width of the next suc- 
 ceeding, since the bar is turned one-quarter over, after each 
 passage through the rolls. The finishing pass, which has 
 the form of the polygon, may be placed similarly either to 
 No. 4 or to No. 5 in the figure. 
 
47 
 
 In the case of light bars of polygon form, one shaping 
 pass is sufficient, which should have nearly the required form ; 
 but the width of this pass should be greater than its height, 
 in order that a sufficient draw may be afterwards applied ia 
 the finishing pass. 
 
 The draw of these passes is proportioned according to the 
 general rules which have already been laid down for the 
 purpose. 
 
 The rolls drawn in Fig. 36, Plate III., contain merely the 
 finishing passes for octagonal bars; these passes are not 
 intended to be used successively in the order given, but each 
 pass is intended to finish a bar of octagonal form, but of 
 different weight. The details of these rolls are, like those 
 described in § 17, not very clear, on account of the small 
 scale. 
 
 § 20. Shapes include a very great and perhaps almost un- 
 bounded variety of passes, the construction of which varies 
 of course with their form. 
 
 In recent times there have been so many, and such 
 quite new apphcations of iron to buildings and machiaery, 
 that the number of shapes has increased enormously. 
 
 Their number is, indeed, so great that it is impossible to 
 treat of them exhaustively, or, indeed, to arrange them satis- 
 factorily, according to a few generic forms. It is, therefore, 
 necessary to be guided by the practical importance of the 
 chief forms, in the selection of so many of them as it is 
 desirable to describe in detail. By handling, thus, first and in 
 greatest detail, the forms which occur most frequently in 
 general practice, a far greater amount of useful information 
 may be imparted, than would be possible were the book to 
 be unduly fiUed up with the discussion of shapes which are 
 sometimes exceedingly difficult to construct, are never pro- 
 duced in large quantity, and are never likely to become a 
 lucrative product of a mill. 
 
 The most important and frequently occurring forms are 
 naturally rails, tyres, angle iron, deck or T, and I beams, 
 some half round forms, such as felloe iron, etc., and finally some 
 of the principal varieties of spike and sash iron ; these will be 
 
48 
 
 taken up respectively in the above order. But before the 
 rolls for the individual kinds are described, it will be neces- 
 sary to preface some general rules for the proper construc- 
 tion of rolls and passes. The composition, size, and form of 
 the various piles for roUing must, therefore, be first consid- 
 ered, as these details stand in intimate connection with the 
 system of passes used. 
 
 § 21. A large original section of the pile contributes quite 
 essentially to the good quality of the interior of the finished 
 product, as well as to its handsome appearance ; but the 
 greater the original sectional area, the greater are the costs 
 of manufacture. Therefore, economical reasons render it 
 necessary that certain limits should be set to the size of the 
 pile. The larger the section of the finished product, the less 
 it is possible to enlarge the original section. The billet out 
 of which wire rod -^-^ inch thick is rolled has a section over a 
 hundred times as large as that of the rod, while the piles 
 which are made up for raUs, tyres, I beams, etc., have often 
 scarcely five times the sectional area of the finished product. 
 
 When these large shapes are heavy, and the reduction of 
 area must consequently be smaU. it is necessary, especially 
 if the form is somewhat complicated, to give the pile a spe- 
 20 cial form correspond- 
 
 ing to the finished 
 shape. This is done 
 in order to employ 
 fewer shaping pass- 
 es. In the adjacent 
 Fig. 20, A and B rep- 
 resent two piles of 
 this kind for an I beam 12 to 18 inches deep, and of the form 
 represented by C. 
 
 It is also necessary to consider the quality of iron de- 
 manded by the individual parts of the pile, and in putting 
 the latter together, that special goodness or quality be pres- 
 ent in any part where the rolling or subsequent use makes a 
 special quality desirable, and, on the contrary, that a cheaper 
 iron of lower quahty be used for other parts, so that the cost 
 
49 
 
 of the product may be as small as possible. Those ^parts in 
 the above figure which are marked a and h are made of iron 
 which has been once or twice reheated and rolled (usually 
 designated as Nos. 1, 2, 3, or " best," "best best," " best best 
 best," according to the whole number of times the iron has 
 been rolled). The parts c, on the other hand, are made of 
 puddle bar. 
 
 Fig. 21 represents another feature of the formation of a 
 pile, which, though not exactly made necessary by the passes 
 used, still affects materiall}'- the economy of their working. 
 
 Fig. 21. 
 
 The section A of the figure is nearly square ; the side view 
 B shows, on the contrary, that the top and bottom bars are 
 considerably longer than the middle ones. This is on ac- 
 count of the fact that, were the bars equally long, the middle 
 part a of the finished beam c would be longer than either 
 flange, and would have to be sawn off to make a beam both 
 ends of which were of the proper form. The pile is, there- 
 fore, arranged as above, in order to economize material. The 
 hatched parts are, further, No. 2 iron, while the rest is pud- 
 dle bar. 
 
 Even the direction of the fibre of individual parts of the 
 pile must be considered in arranging the latter. For in- 
 stance, when single-lipped chairs were rolled at 
 Zoptau, in Moravia, to the form A, Fig. 22, the 
 individual chairs being sawed off and the lips 
 bent down simply, as shown in B, very many 
 lips or feathers were broken off, so long as all 
 parts of the pile were arranged as usual, and so 
 that the fibres all ran in the direction of the 
 length of the pile, while the feathers were bent 
 across this direction. This difficulty was obvi- 
 ated by taking out the longitudinal bar from the 
 
 Fig. 22. 
 A , 
 
50 
 
 position a, and laying it in its place and next eacli other nar- 
 row strips 4 inches wide, with the fibre running crosswise. 
 These strips were cut from No. 3 bars, while the other parts, 
 with the exception of the three layers of puddle bar, h, h, h, 
 consisted of No. 2 iron. 
 
 It must also be here remarked that very large shapes are 
 not usually rolled out of a single pile, but are composed of 
 parts rolled separately, and afterwards welded together. 
 Thus, for girders of a depth of two feet or over, the top and 
 bottom flanges, a and b, are rolled separately, while a wide 
 bar or plate, c, forms the web, the respective pieces being 
 joined by the double channel bar, m. The flanges are 
 usually heavy deck beams (T iron), of a strength 
 ' • corresponding to the size of the desired girder. 
 The parts are fitted to each other, as shown in 
 the adjacent Fig. 23, are held together by clamps 
 and screws, and are then welded, piece by piece, 
 along both lines of junction at once. The dis- 
 * tance welded at each heat is about 1 foot. The 
 girder is supported on rollers, so that it may be 
 moved in the direction of its length. At right 
 angles to this line of rollers a small raihoad is 
 laid down, just beside which latter is placed a 
 pecuhar anvil. Two small blacksmith's fires, 
 with several tuyeres, are borne on the railroad, 
 one on each side of the girder, so that they may 
 be pushed together under it, or apart, and away from it. 
 When by their use a welding heat has been got up on both 
 sides of the girder,. they are pushed back, and the part 
 heated is rolled upon the anvil, when it is struck and welded 
 by two 40-pound hammers, worked from a shaft running 
 crosswise underneath them. To insure a good weld the 
 channels in the double-channel bars need not be more than 
 1^ to ^ of an inch deep. 
 
 § 22. In the case of all passes described up to § 19, inclu- 
 sive, the construction could be, and actually was, so carried 
 
51 
 
 out that the pressure was entirely equal toward each side. 
 The bar had thus no tendency to bend sidewise ; the diame- 
 ter of the grooves was, however, so proportioned that the bar 
 might be bent slightly downward against the guards. As far 
 as possible, the same rule is observed in rolls for shapes ; 
 and if a shape may be divided vertically by the centre line 
 into two equal parts, the observance of this rule occasions 
 no difficulty, and the relation holds good that ordinarily ob- 
 tains between the draw and the spread of common bar iron. 
 The only difference between the cases is that in the latter the 
 draw is uniform across the width of the bar, while in the for- 
 mer it must necessarily be quite different, in order to form 
 the required shape. The question naturally arises, how great 
 may the difference of draw be allowed to become ? 
 
 If no draw was applied to some parts of the bar, the form 
 desired would be very quickly attained by the use of a few 
 passes. If, for instance, the attempt was 
 made to roU the channel iron in Fig. 24 
 simply by compressing the middle part, ^^"~ZZ'£'ZZ''^ M 
 a or 6 6, of a flat bar, the result would be 
 that the sides or flanges, c c or dd, would ^ ^ 
 also be drawn out — not so much, indeed, ^^^^^^^^^J 
 as the middle or web, but more or less ^M.lZZ..i.Z.Z.^M 
 according to the quahty of the iron, and 
 would, therefore, become ragged, or, even in the case of the 
 very best iron, would at least lose their proper form. There- 
 fore, in order to preserve the proper shape, it is necessary to 
 draw the flanges also, to fill out the whole pass thoroughly 
 with iron, and to use for the flanges a specially good iron, 
 which has been reheated and reroUed, or, in other words, a 
 No. 2 iron. 
 
 The permissible differences in draw depend, on the one 
 hand, upon the absolute area of the section of the pass, and, 
 on the other, upon the quality of iron. In the case of large 
 deep passes, when the greatest reduction upon the upper, 
 as well as upon the lower side of the bar, amounts to about 
 1 inch, the smallest reduction may be restricted to or of 
 
52 
 
 an incli, as shown on Plate VIII., in Fig. 82, in full size 
 passes for girders, which are used at Eeschitza, and also in 
 the passes for channel bar and girder iron, shown in Figs. 
 86 and 87, on the same plate. The difference between the 
 draws or pressure upon the different parts becomes less in 
 those passes which precede the finishing passes, while in the 
 latter, in which it is necessary to complete the form of the 
 flanges, these are drawn out rather more than any other 
 part. The poorer the quality and the greater the hardness 
 of the iron, the more nearly uniform must be the reduction 
 of the different parts of the section. 
 
 In the case of channel bars, with very deep channels, or 
 in the case of V iron, the rolUng can be rendered very much 
 easier if the shaping passes are so propor- 
 tioned that the ridges (a a) are formed ; 
 in the finishing pass these ridges are 
 much more easUy pressed into the flanges 
 or sides (6 6 in A, Fig. 25) than the su- 
 perfluous iron of the middle part would be. 
 The iron from the ridges (a, a) can be 
 pressed down into the sides (&, h) without 
 materially lengthening or drawing out the 
 bar, but it would require a very heavy draw to force the 
 iron out of the centre (c) into the sides. The arrangement of 
 passes for iron of the above form, is shown in Figs. 93 and 94, 
 on Plate IX., and the principle is susceptible of very 
 general application in rolling shapes. 
 
 In every shape in which the iron does not lie equally on 
 each side of the vertical centre line, as, for instance, almost all 
 forms of tyres, rails, angle iron, with flanges of unequal width, 
 and many other forms, the draw upon the two sides must 
 necessarily be unequal. So, for instance, when the first T, 
 or flanged rails were rolled at Frantschach, in Carinthia, 
 they were rolled in the passes shown in full size by Fig. 50, 
 on Plate IV., which were closed passes, as is usual for 
 shapes, and for which the bars were turned 180 degrees, or 
 half over, at each pass. The rolHng itself was in this case 
 
63 
 
 quite easy, and tlie desired form perfectly attained ; but the 
 rolls themselves were very soon worn out, because they 
 were considerably weakened by the necessary height of the 
 last roughing and first shaping passes, but chiefly because 
 the draw was so unequal that the collars were pressed ob- 
 liquely against each other, and were abraded to such an 
 extent that the rolls became useless. 
 
 The arrangement of passes, which is now used in rolling 
 rails, as well as shapes generally, consists in the use of 
 one or more flatting passes, as represented, for instance, in 
 the rail rolls shown in Figs. 44, 45, and 46, on Plate lY. In 
 these passes the whole mass of iron is somewhat compressed, 
 but the principal pressure comes upon the flanges, to com- 
 press and spread them out. A comparison of these passes, 
 as shown in full size in Figs. 48 and 49, with the older forms 
 represented in Fig. 50, show quite plainly to how great an 
 extent the former obviate the unequal draw. 
 
 These inequalities of draw may also be obviated in some 
 cases by another method, which is, however, less generally 
 known and iised, but which is especially useful in rolling 
 angle iron with flanges of unequal width. It consists in so 
 proportioning the first shaping passes, as shown in Figs. 
 61 and 62, on Plate V., that the narrower flange remains the 
 thickest, so that in the finishing passes it is necessary to use 
 a heavier draw upon this side, or, in other words, to exert more 
 pressure upon its surface than upon that of the wider side. 
 Thus the total pressures become nearly equal, since the pres- 
 sure upon the wider flange is less, in about the same proportion 
 that its superficial area is greater than that of the narrower 
 flange. By this means the unequal wear of the coUars is very 
 much lessened. The unequal draw or pressure might be ob- 
 viated by varying the angle of the flanges in the successive 
 passes, so that the final angle would be the one required ; 
 but such an arrangement would entail many disadvantages, 
 and is, so far as I know, nowhere in use. 
 
 Finally a third method, which has been long in use, con- 
 sists in so arranging the journals of the rolls (which should 
 
64 
 
 in this case be brasses), that they receive upon themselves 
 
 „. „- the side-thrust ex- 
 
 Fig. 26. 
 
 erted on the roll. 
 
 ■ ■S, !i T T '] arrangement 
 -j'7^pf;i_— r r-JfrTT^' — is shown in Fig. 
 ^^^£^^^^ ^ ^^ ^4::^^^ ^^^^^ 
 
 ^'^^^^ dotted lines, and as 
 
 l^. _---.._J if it had been push- 
 
 ed forward several 
 
 inches in the direction of its axis, whereas it presses during 
 use quite firmly against the brass, c. The brasses have a 
 strong flange or collar on the roll side, between which collar 
 and the cast-iron box two slender steel wedges are driven 
 down vertically. The journal-box does not press against the 
 two set screws (s, s), as usual, but is pressed directly against 
 the chock (a, a). When the steel wedges are driven down, the 
 brasses are drawn out of the box and pretty firmly forced 
 against the collar of the roll. It is therefore possible, by 
 this arrangement, to relieve the collar from the side-thrust 
 occasioned by unequal draw, and transfer the thrust to the 
 brasses. The latter must not, of course, be wedged too firmly 
 against the roll, as the strength of the neck would thereby 
 be too much taxed, and the friction greatly increased. 
 
 When it is not possible to avoid an unequal draw, the bar 
 has a tendency, at its exit from the pass, to bend toward the 
 side where the draw has been least. But the bar must be 
 straight in order that the rolling may be continued ; it is there- 
 fore necessary to fasten upon the bearing bar proper, blocks, 
 called guides, at the exit end of the pass, so that the bar may 
 be straightened as it passes out between them. Sometimes 
 gimrds must be used upon the top roll to prevent the bar from 
 bending upward. 
 
 If a very heavy draw must be applied in the finishing pass, 
 the bar is considerably widened or spread, as in the case of 
 an intermittent pass for spike iron, the product of which is 
 shown in Fig. 27. When the part a is greatly reduced to 
 
55 
 
 form the head of the spike, it is very much widened, and if 
 the pass does not freely per- 
 mit this spread, the forma- 
 tion of fins, as shown in sec- s_ 
 tion at 5, is unavoidable, and 
 the bar is spoiled. In order 
 to prevent this the pass be- i ' 
 fore the last must not only be 
 of somewhat less width than | J 
 the latter, but it must roll 
 the iron in wedge form, as shown in section by c, with bevel- 
 led edges upon the side which is periodically compressed by 
 the finishing pass. Seen from above, the spike rod has the 
 form shown by B, The formation of fins is aided by the cold 
 state in which the bar is rolled through the last pass, as cold 
 iron spreads more than hot iron under the same draw. Fur- 
 ther, cold bars wear out intermittent passes very quickly, and 
 therefore, for this double reason, the iron must be rolled 
 through them as hot as possible. 
 
 Passes for shapes are, as a rule, closed passes, and as this 
 variety guides the bar pretty well by itself {vid. § 3, 2), other 
 accurate guides fitted to the special form are the less neces- 
 sary, since the increase of width in successive passes is in- 
 considerable. But this circumstance, although it renders 
 the guidance of the bar more certain, also makes it neces- 
 sary that the latter should be correctly introduced into the 
 pass, in order that the edges of the former, which closes the 
 pass, may not begin to cut as they take hold. Hence it is 
 necessary to place suitable guides upon the bearing bar in 
 front of the rolls, in order that the bar may be introduced 
 straight and fairly into the pass. 
 
 One circumstance yet remains to be noticed, viz., that for 
 all complicated passes for shapes it is very important to 
 keep the guards sharp and in proper position, and to cool off 
 the passes with a great deal of water, especially in the case 
 of those in which the bar is likely to stick, and to be imme- 
 diately wrapped round the roll. 
 
 Fig. 27. 
 
 A 
 
 5 
 
56 
 
 § 23. It is evident from the foregoing (under reference to 
 § 7), that the composition, size, and form of the pile or other 
 material, affects the shape and draw of the required passes ; 
 these elements are also essential in determining the requisite 
 diameter of the rolls. The size of the rolls is, indeed, a ques- 
 tion of almost as much importance as the form of their 
 passes ; if the former are small, they frequently break and 
 suffer great wear, while if they are, on the other hand, large, 
 they are more expensive, heavier to handle, and increase the 
 lift to get the bar over or through (a three high mill) to the 
 front. It will be well to consider the characteristics of each 
 kind. 
 
 Small rolls do not require very heavy housings or fittmgs, 
 are quite easily handled, contain a comparatively small weight 
 of metal, and the bar need not be lifted very high to get it 
 over them. They must, however, be pretty long to contain 
 the necessary passes, while some of them are cut very deeply 
 into the bodies; they are, therefore, comparatively weak. 
 They require a pretty high speed of the engine, if the latter 
 is direct-acting, or if it is not, occasion great wear and tear of 
 gearing ; while the grooves, owing to their small circumference, 
 wear more for the same length of bar than would the grooves 
 of a large roll, and consequently sooner require redressing. 
 They resemble in their action a ^hammer with narrow dies 
 and draw out quickly, with little spread; but the draw 
 operates more quickly than it would between large rolls, on 
 account of the relative shortness of the wedge-shaped space 
 which the bar must traverse to reach the normal section of 
 the pass. Since the bars are, therefore, as it were, less pre- 
 pared for the normal draw of each pass, all the strains be- 
 come more sudden than would be the case in larger rolls. 
 A small roll limits the draw, since the latter, if too great, 
 would break the roll, and this limitation of draw increases 
 the number of passes, other circumstances being similar and 
 equal ; if, however, the draw remains the same, the rolls must 
 be shorter. For instance, a case of this kind occurs to me 
 in which the rolls of a three high mill broke down under the 
 work for which they were designed; the housmgs could 
 
57 
 
 neither be made stronger and larger to receive heavier rolls, 
 nor could they be thrown aside ; consequently the only alter- 
 native was to shorten the existing rolls and add one whole 
 set to receive the passes lost by shortening the rest. An- 
 other feature of small rolls is that those passes in which the 
 draw is at all heavy, require deep notching to insure their 
 hold ; but if these notches are pretty deep and are not very 
 carefully considered in the formation of subsequent passes, 
 they are apt to leave unsightly marks or laps on the finished 
 product. 
 
 Large rolls are, of course, more expensive than small ones, 
 require heavier housings and fittings, and give rise to great- 
 er friction. They have, however, great and manifest advan- 
 tages over small ones. They are comparatively strong, since 
 with the same length their strength increases as the square 
 of the depth or diameter ; 21-inch rolls being thus over ^ 
 stronger than 18-inch, while the increase of diameter is only 
 ■g ; the rods will therefore stand a heavier draw and still re- 
 tain proportionately greater strength. The " nip" of large 
 rolls is far better than that of small ones, on account of the 
 greater surface working on the bar, and consequently greater 
 friction ; the notches may therefore be made leas deep, there- 
 by preserving the surface of the finished bar. In order to 
 get the best efibct or work from large rolls, the draw should 
 be slightly heavier than in the small ones, which they would 
 replace, the speed at the circumference of the respective rolls 
 being the same. The large rolls will then do their work as 
 quickly as small ones, while the iron is not strained so much 
 as in the latter, by reason of the more gradual wedge-shaped 
 approach to the normal section of the pass. In order that 
 the circumference of a 21-inch roll might have the same speed 
 as that of an 18-inch, the engine, if direct-acting, might run 
 about ^ slower, that is, about 56 revolutions instead of 65, and 
 this decreased speed would increase the endurance of all parts 
 of the rolls, by lessening all shocks and rendering them less 
 sudden. For these reasons, as well as much greater certainty 
 against unprofitable stoppage of work through the breaking 
 down of rolls, the modern practice has been to use very heavy 
 rolls, and this practice is aheady very general. 
 
58 
 
 These remarks on the characteristics of large and small 
 rolls are, of course, merely comparative ; the speed, draft, and 
 form of pass required for any particular iron must be deter- 
 mined separately by very careful experiment. 
 
 It is very important in all rolls to avoid fins of any kind, 
 and, indeed, the methods of rolling vtdthout forming fins make 
 up the most important part of roll-turning, since with them 
 are connected all questions of draw and shape of passes 
 while the quality of stock which can be worked up in a mill, 
 depends very materially on the construction of its rolls. The 
 reader wiU, therefore, do well to notice the various methods 
 of preventing fins pursued at different works, as shown in 
 detail hereafter, as there are no details in which skilful de- 
 sign is better displayed than in these, while the comparative 
 study of these methods will impart a very good knowledge 
 of many important principles. 
 
 § 24. Let us now consider in detail the rolls for the most 
 usual shapes, beginning with those for rails. The first ex- 
 ample of such rolls is given in Plate III., in Figs. 39 and 40, 
 in tV full size, as used in a Westphalian mill. The passes of 
 these rolls are represented in full size in Figs. 41 and 42. 
 The section is one which was at one time quite common. In 
 these rolls the bar must be rolled through 10 passes, which 
 are equally divided upon two pairs of rolls, viz., the rough- 
 ing and the finishing. The latter contain, however, 6 passes, 
 as the finishing pass is duplicated ; the second pass being, of 
 course, used only after the first has become too rough. It 
 seldom occurs that rails of this size are rolled through only 
 10 passes. This is rendered possible by the proportionally 
 narrow flange, as only one flatting pass (the first in the 
 finishing rolls) is required for forming the latter. 
 
 It will be remarked that one edge of the flange is made 
 round, while the other is of the proper shape. This peculi- 
 arity is owing to the necessity of preventing fins, which 
 would be formed between the former and the collar, were 
 not some special means adopted to prevent their formation. 
 The most effectual means to this end has been found to be 
 
59 
 
 the leaving away of tlie metal (in the part where a fin is to be 
 feared) in such a manner that when the flange or other part 
 comes into the pass, the draw of the latter may only fill out 
 the pass well, instead of forcing the metal into all its crev- 
 ices. This principle has been long successfully applied in 
 several different ways, as will be shortly pointed out. A fur- 
 ther peculiarity of these rolls, is the fact that the division of 
 the roughing passes is made a considerable way above the 
 pitch line. This method allows the fins formed in one pass 
 to be thoroughly rolled out in the next succeeding pass, since 
 the bar is so turned over as to brmg the fin to the lower 
 part of the succeeding pass, where the bottom forces the fin 
 back into the bar. This arrangement also prevents, to a 
 great extent, the formation of fins at any formed arc at once 
 sheared off. The fin would not be rolled out if the pass divi- 
 sions were at the pitch line, unless some such arrangement 
 was used as that adopted by the Messrs. Fritz, viz., forming a 
 deep groove at those parts which come opposite to any division 
 or opening. This plan prevents fins, and is very convenient. 
 
 By means of the flatting pass a special advantage is ob- 
 tained, in addition to the preservation of the rolls, mentioned 
 in § 22, namely, that the work upon the flange is much, 
 greater than that upon the head, while the contrary is the 
 case in the older arrangement, shown in Fig. 50, on Plate IV. 
 The greater amount of work renders the iron in the flange 
 more fibrous than that in the head, which generally shows 
 some grain. It thus becomes easier to fulfil the requirements 
 of many engineers, that the flange shall be fibrous and the 
 head hard, or *' fine grain," though this difference is in fact 
 produced, in a marked degree, only by employing, for the 
 various parts, materials of different hardness. 
 
 Two shaping passes are turned upon the roughing rolls in 
 Fig. 39, which, like the welding passes, are turned as open 
 passes. In order that the rolls may not be moved sidewise 
 during the action of the shaping passes, the end collars, as well 
 as several of the body collars, are so turned that they fit toge- 
 ther conically ; this construction naturally aids very much in 
 keeping the rolls in their proper position. In the flatting pass 
 
60 
 
 the under part of the bar (the future head of the rail) is 
 reduced scarcely at all, while the partially formed flange 
 is quite forcibly spread out. The less the under part of the 
 bar is reduced, the more easily and thoroughly is the flange 
 spread out ; but in this case the iron must be very tough, or 
 the flange is apt to crack. The last passes of the finishing 
 rolls have the peculiarity of being " half open" passes 
 {vide § 3) ; the sides of the head are thereby much more per- 
 fectly rounded and finished than would be the case in an 
 open pass. 
 
 Figs. 57 and 58, on Plate Y., represent the passes of the 
 rail train of the mill at Gratz, which belongs to the " Siid- 
 bahn," or Southern Railroad. In these rolls 12 passes are 
 used, 6 of which are on one pair and 6 on the other pair of 
 rolls. By reference to the roll drawings in Figs. 39 and 40, 
 on Plate III., it is rendered easy to draw the rolls which 
 belong to these passes, and which were omitted to save room 
 (see also § 13). The flange of these rails is considerably 
 wider than that of the former section ; hence two flatting 
 passes are used, viz., pass 6 in the first and pass 3 in the 
 second pair ; thus the flange is spread out more gradually to 
 the necessary width. Pass 3 of the finishing rolls shows at 
 a a construction the principle of which is similar to that of the 
 similar pass of the Westphalian roUs previously mentioned, 
 but in this case differently carried out, with the special pur- 
 pose of so shaping one edge of the flange at a that it will be 
 rolled down again in pass 4 by the former, thereby prevent- 
 ing the formation of a fin at a V which purpose is also per- 
 fectly accomplished in rolling. 
 
 An example of a rail train with many passes is given on 
 Plate IV., in the Figures 44, 45, 46, the corresponding passes 
 being represented in full size in Figs. 47, 48, and 49. 
 This train is in use at Pravali, in Carinthia, for rolling Besse- 
 mer steel, and it rolled the first rails of that kind successfully 
 made in that part of Austria. Between the three pairs of 
 rolls belonging to the train 15 passes are evenly divided and 
 are used successively ; but the actual number of times the 
 bar is passed through the rolls is about 20, in two heats. 
 
61 
 
 The reason for sucli a large number of passes is not at all 
 the difficult section of rail, but lies altogether in the physical 
 characteristics of Bessemer steel, as being comparatively hard 
 and resistant. 
 
 The somewhat pyramidal ingots of Bessemer steel are of 
 square section, 7 to 8 inches on a side, with rounded corners. 
 They are principally cast of steel, containing 0.3 and 0.5 per 
 cent, of carbon, as that containing over 0.5 per cent, would 
 not only not be so easily rolled, but would scarcely be tough 
 enough to stand the required deflection tests, because they 
 would be comparatively hard and stiff. The rough ingots 
 are heated to a full yellow heat, and are then rolled three 
 times through the first pass, twice through the second, and 
 twice through the third — therefore seven times through the 
 first three passes ; finally, the bar goes twice through the 
 first shaping pass, and is then immediately returned to the 
 furnace for the wash-heat, after having in the heat gone 
 nine times through the rolls. The rolls are opened about 
 half an inch for the first passage of the ingot, and are grad- 
 ually closed again ; the ingot is turned quarter over (90°) at 
 each passage through the first three passes, but half over 
 (180°) each time in the fourth. The second heat is higher, 
 nearly approaching a white heat, and in this heat the ingot 
 goes once through each of the remaining 11 passes, and is 
 thus passed 20 times through the rolls. Out of the 11 passes 
 used in the second heat, the second, fifth and eighth are flat- 
 ting passes. 
 
 It is interesting to know that these rails were 21 feet 
 (21.77 Eng.) long, and weighed 368 (454.33 Eng.) Viennese 
 pounds. The ingots weighed 415 to 430 (513.35 to 530.87 
 Eng.) pounds in the rough. The loss in heating and rolling 
 amounted to 4.9 per cent. Seven heats of four (4) ingots each 
 were made in one furnace during the turn of 12 hours. The 
 percentage of imperfect rails varied between 5 and 10 per cent. 
 
 In Figs. 39 and 44, which represent the roughing rolls 
 of two rail trains, and Figs. 41 and 47, which represent 
 the respective passes in full size, it will be noticed that the 
 openings between the body-fiUets are some way above the 
 
62 
 
 ■pitcjb liiie. - This is so arranged designedly, in order that fins 
 may not be easily formed in the passes used in both cases. 
 This method insures that any fins formed are well rolled in, as 
 well as prevents their formation. That is, the bar is turned 
 half over at each pass, and a fresh surface comes before the 
 opening of the pass ; the resulting fin is thoroughly rolled in at 
 the bottom of the next pass, and in the third pass a straight 
 surface (on which no fin has yet been formed) comes before 
 the opening; that part of the bar on which the fin was formed 
 in the first pass coming a good way below the parting of the 
 third. The flaring sides of the body collars furnish additional 
 protection against fins (vide end of § 30), while the surface of 
 the bar is kept smooth and sound, because no part where a 
 fin has once been comes again into a position where another 
 is likely to be formed. 
 
 Fig. 37, on Plate III., represents a pair of rolls used in a 
 Silesian mill for rolling mine rails ; the corresponding passes 
 are shown in full size in Fig. 38. In this pair there are 6 
 passes intended for successive use ; the seventh is a spare or 
 duplicate finishing pass. This pair contains only finishing 
 passes ; the roughing rolls are omitted, because they cor- 
 respond exactly with the Gothic roughing rolls for bar-iron 
 (vide § 11 and Figs. 5 and 6 on Plate I.), as is evident from 
 the form of the first finishing pass. In these rolls for such 
 a small section a flatting pass is also used, both to preserve 
 the rolls from undue side-thrust, and to make the flange com- 
 paratively tough and soft, as discussed above in detail, 
 
 § 25. It is necessary to work up old rails, imperfect rails 
 or wasters, and crop ends, in the formation of the rail pile ; 
 it will therefore be useful to consider the best means of roll- 
 ing them into the proper shape. 
 
 It was formerly, and is now to some extent, customary to 
 cut the rails to the length of the pile, and to use one or two 
 such cut pieces in the formation of the latter. These pieces, 
 on account of their irregular form, did not fit well into an 
 ordinary pile of puddle and reroUed bars (tops and bottoms) ; 
 therefore, in order to avoid interstices between the bars and 
 
rails, it was necessary to roll filling pieces of special, '-foriJi 
 adapted to fit the ran more or less closely. TKe use of rail 
 
 ends or pieces is now almost en- 
 tirely restricted to the pile for 
 tops and bottoms, and it has been 
 ^^^^^ found more advantageous to roll 
 ^^^^^ the pieces down singly into a flat 
 ^^^^^ bar. For this purpose old rails 
 and wasters are commonly cut 
 cold into lengths of 5 or 6 feet, reheated and rolled into flat 
 bars twice or thrice as long as the rail pile. These bars may 
 be made in form and size similar to puddle bars, and may be 
 conveniently used in any desired proportion in any part of 
 the pile. 
 
 The crop ends are usually rolled down immediately after 
 cutting. They vary, of course, considerably in length, but 
 may be used in the pile together with the ends cut off the 
 puddle bars. 
 
 Fig. 53, on Plate V., shows a pair of rolls adapted to rolling 
 crop ends, etc., into flat bars ; they are in use at the Horst 
 mill near Steele, and produce bars of 3 to 4 inches in width. 
 The corresponding passes are shown in full size by Figs. 54, 
 55, and 56. The two first passes reduce the height of the 
 rail about | of an inch (drawing it out of course), but are 
 principally useful in bending the flanges back upon the web, 
 as shown by the dotted lines in the first pass. Between the 
 second and third passes the rail must be turned one-quarter 
 over, and half over for the fourth as well as for the fifth. 
 The fifth pass turns out a flat bar, 4 inches wide by 1 inch 
 thick. If the bar is desired 3 inches wide, it must be turned 
 quarter over (or upon its edge) for the flatting pass No. 6, in 
 which its width is reduced from 4 to 2U inches; it is then 
 turned again one-quarter over, and rolled through No. 7 into 
 a bar 3 inches wide by 1 inch thick. 
 
 Those rails— which are composed of different materials in 
 the flange, web, and head— are seldom rolled out into a bar 
 of this kiud, but are usually cut by means of shtting roUs into 
 three parts— viz., head, web, and flange — which may then be 
 used again in accordance with their respective qualities. 
 
64 
 
 By the use of rolls so constructed as to gradually press 
 the metal in the flange and head into the web, flat bars may 
 be rolled from old rails and crop ends, without bending the 
 flange at all over upon the web. This method makes a 
 smoother bar, while it obviates the weld of the flange upon 
 the web, and hence materially reduces the whole number of 
 welds in the pile. 
 
 § 26. A pair of tyre rolls used in a SUesian mill are shown 
 in Fig. 68, on Plate VI. ; they contain 6 passes, which are 
 given in full size in Fig. 69. Figs. 70 and 71 also represent 
 tyre rolls in use at the same mill, containing 6 passes on two 
 pair of rolls ; Figs. 72 and 73 exhibit these latter passes in 
 full size. In one aspect it is certainly simpler and cheaper 
 that one pair of rolls should contain the 6 shaping and finish- 
 ing passes, and such is generally the case. It often hap- 
 pens, however, that small differences in form and size of 
 tyres are frequently demanded, and it is then much better 
 that the four shaping passes should be turned on one pair 
 of rolls, which can then remain in their housings, while the 
 small pair, with the two finishing passes, may be easily and 
 quickly changed. In order that these small alterations of 
 section may be readily accomplished, the last pass but one 
 is made a flatting pass, and turned, of course, upon the 
 short rolls ; while the flatting pass in Fig. 68 is the third to 
 be used. 
 
 By means of the flatting pass the unequal side-thrust of 
 the rolls is somewhat diminished, as mentioned in § 22, and 
 also especially discussed with reference to rail rolls. The 
 chief object of their use is, however, the more perfect for- 
 mation of, and a thorough work upon the flange of the tyre. 
 
 The above rolls require in use a pair of roughing rolls, 
 with the necessary passes. These are omitted here, but 
 their form and draw may be readily derived from the next 
 following pair of tyre rolls. 
 
 Fig. 74, on Plate VI., represents, in full size, the welding, 
 drawing, shaping, and flnishing passes used in a Styrian mill 
 for rolling tyres. Ten passes are used, as the figure shows. 
 
65 
 
 Of these, tlie first five,wliicli are box passes, are on the rough- 
 ing rolls ; while the last five, on the other hand, are turned 
 as closed passes upon the finishing rolls. Of the five rough- 
 ing passes, the first three are to be considered as welding, 
 and the last two as drawing passes. The bar must be turned 
 one-quarter over before entering the 2nd, 3rd, and 5th passes 
 respectively, but half over before entering the 4th. The 
 third of the five finishing passes is a flatting pass. 
 
 A noteworthy difference exists between the Styrian and 
 Silesian passes, viz., that the former tyres are finished con- 
 cave on the inner side, while the latter are left straight. 
 The object of making the inner side concave is that it may 
 become straight after the tyre is bent into form ; for if it 
 were left straight it would be found to be rendered quite con- 
 vex, which convexity, or bulge, must afterward be turned offj 
 thereby increasing the cost of labor, and causing a greater 
 waste of metal. Something similar occurs with reference to 
 the other dimensions of the tyre : the inner side not only 
 bulges out, but also becomes wider on account of the bend- 
 ing, while the outer side, or tread, becomes narrower. 
 Therefore, in order that the other two sides of the tyre may 
 remain parallel, it is necessary to roll the bar in such a 
 manner that the inner side shall be finished narrower than 
 the tread ; while in the rolls shown in Figs. 68, 71, and 74, 
 exactly the reverse is the case, as the inner side must be 
 finished wider than the tread. In order to insure that the 
 inner side be narrower than the tread, it would merely be 
 necessary to turn the finishing pass half over, thus bringing 
 the tread uppermost, by which means it — viz., the whole tread 
 side or outside of the tyre — might be readily made wider as 
 desired, being flatted out somewhat by the former of the 
 upper roll. This method is applied in some mills, but in 
 the majority this self-evident advantage is neglected, in 
 order to finish the flange somewhat more accurately in the 
 lower roll. 
 
 Tyres for disc wheels must be rolled with a rib or fillet on 
 the inner side, to which the disc, which takes the place of 
 spokes, may be fastened. Fig. 43, on Plate III., represents, 
 
66 
 
 in full size, the shaping and finishing passes used in a 
 Silesian mill for producing such tyres. 
 
 These examples of rolls for tyres are adduced rather to 
 show an interesting and useful series of pass constructions 
 than as anything at present to be recommended. Iron tyres 
 are now seldom used, and steel tyres are almost universally 
 weldless, and rolled in a peculiar form of mill, some of the 
 rolls of which will be described in § 32. 
 
 § 27. There are two circumstances to be considered in 
 draughting passes for angle iron, viz., the equal or unequal 
 length of the sides, and the absolute area of the section. 
 When the sides are of equal width, angle iron is much more 
 easily and cheaply rolled than when they are not, as in the 
 latter case the rolls soon become worn out by the irregular 
 side-thrust. The latter difficulty is materially diminished 
 when the shorter side is kept thicker than the longer, as 
 described in § 22. 
 
 If the sides of the angle iron are equal and plain, and if 
 their edges are not required to be very smooth, a single pair 
 of rolls may be used to produce iron varying both in width 
 of the sides and in the thickness of the metal. The passes 
 of such rolls must be constructed for the greatest desired 
 width of side, as when a section with narrower sides is to be 
 rolled, it is sufficient to give it the required size in the last 
 pass, since it is not necessary that the iron in the sides 
 should entirely fill out the grooves of the shaping passes. 
 This method resembles that employed when flat bars are 
 rolled between plain roUs, viz., rolls similar to plate-rolls. In 
 the latter case, however, the width of the bar varies some- 
 what, and the corners are not sharply rectangular. The only 
 difference between the two cases cited above is to be ob- 
 served in the fact, that in rolling angle iron great care is 
 required to introduce the centre of the angle exactly in the 
 centre of the pass ; for if the bar should deviate at all to the 
 right or left, the width of the sides would vary considerably. 
 In order to insure the proper introduction of the bar without 
 using accurate guides, it is only necessary to turn the passes 
 
67 
 
 upside down, as it were, as shown in Fig. 65, on Plate V. ; 
 that is to say, that the former is placed on the bottom instead 
 of the top roll. In this way it is rendered easy to introduce 
 the angle exactly, by allowing it to rest on the former, and 
 be carried by the same into the pass. If the groove of the 
 pass was turned on the lower roll, the diameter would be so 
 much decreased that it would be difficult to adopt the above 
 method, as the bearing-plate in front of the roll would cut olf 
 a good deal of the circumference of the latter. Further 
 comments on the use of such inverted passes will be found 
 on page 69, in connection with the description of Fig. 60, on 
 Plate V. In order to make the flanges thicker it is, of 
 course, only necessary to separate the rolls as required. 
 
 It is important to consider the absolute sectional area of 
 the finished angle iron, for when the same is small, the bar is 
 generally worked out from a nearly square section, whereas, 
 on the other hand, the larger sections are commonly worked 
 out gradually from a flat bar into the required rectangular 
 form. The first method is the simplest, but requires a 
 greater proportional reduction of the original area than the 
 second method, and is, therefore, not so economical or con- 
 venient for rolling large sections. Fig. 51, on Plate IV., re- 
 presents the shaping of the angular section from the flat 
 form, while Fig. 52 shows the section as drawn out of the 
 square bar or pile. The dotted lines, a, h, c, d, in each fig- 
 ure represent the original section of the pile or bar, while 
 the finished angle iron is designated by the hatched section. 
 In each case five passes are required to finish the bar. 
 These two sets of full size passes, representing both of the 
 above methods, have been many years in use at Frantschach, 
 in Carinthia. 
 
 When the angle iron is rolled up from the flat form, the 
 length of each side may increase, while the distance between 
 the edges remains the same, or even decreases, because the 
 sides are gradually bent in. On this account it is not 
 so necessary, as in the case of other forms, that one pass 
 should be wider than the preceding ; individual passes may 
 indeed be narrower than those preceding them, as shown by 
 the passes for angle iron with unequal sides, which are given 
 
68 
 
 in Figs. 61 and 62. In order to understand how it is pos- 
 sible to introduce the bar into a pass narrower than the one 
 it has just left, it is only necessary to consider that the point 
 of the angle projects most, and is first seized by the rolls, 
 and the new angle is formed before the greater width of the 
 sides can become troublesome, or before the edges begin to 
 be cut. 
 
 For the sake of simplicity and ease in turning the rolls, 
 the deep grooves are, as a rule, aU placed in the bottom roll, 
 although this position has the disadvantage that the bar 
 cannot, as is usually the case, be turned half over, in order 
 to prevent the formation of fins. In order to obviate this 
 disadvantage, more or less, the two last passes are often so 
 constructed {vide Fig. 61, Plate V.), that, though really closed 
 passes, they are yet open at the edge of the side. Some- 
 times, to accomphsh the same object, the passes are so 
 arranged that the bar maybe turned once or twice half over ; 
 this is the case at the Donawitz mill, near Leoben ; while 
 Fig. 59, on Plate Y., represents similar rolls in use at a Sile- 
 sian mill. 
 
 The first formers should be roughened with a chisel, as 
 shown in Fig. 60, in order that they may take a firmer hold 
 of the bar. 
 
 Fig. 61, on Plate V., represents in full size a pair of 
 finishing rolls for heavy angle iron with unequal sides. Fig. 
 62 gives the corresponding passes in fuU size. The rolls 
 are in use at Zeltweg in Styria. The great difference be- 
 tween the diameters of the roll in the various grooves is 
 worthy of remark; these differences are, in this instance, 
 caused by the large size of the section rolled. As a rule, 
 however, all rolls for angle iron present great differences of 
 this kind. Further, Fig. 63 shows passes for medium heavy 
 angle iron with equal sides ; they are shown in their position 
 on the rolls and in full size. After an examination of 
 Fig. 61 it is easy to construct rolls for these passes, espe- 
 cially as the pitch line is marked on the drawings. 
 
 With reference to the roughing rolls belonging to these 
 finishing rolls, it is necessary to consult § 13 and Fig. IP, on 
 
69 
 
 Plate n. ; for an inspection of Figs. 62 and 64, on Plate V., 
 shows at once that large sections must be rolled from corre- 
 spondingly large and heavy flat bars. 
 
 Fig. 60, on Plate Y., represents a pair of finishing rolls for 
 angle iron with sides of equal width, each side being 
 inches wide ; they are in use at Piela in Upper Silesia. 
 There are two especially striking points to be remarked in 
 these rolls ; the first being, that all passes have the same 
 width, viz., 4| inches ; the second, that the iron is indeed 
 rolled out of the flat bar ; but the shaping is effected in one 
 pass, and with nearly the full finished angle. This is a method 
 which the moderate size of the angle iron renders practi- 
 cable, but which would be, even for small sections, seldom 
 advisable or applicable. It is, therefore, unnecessary to 
 represent the individual passes in full size. Fig. 65 repre- 
 sents a set of passes for medium and small angle iron with 
 equal sides, and which is shaped from a square or diamond 
 through the half diamond bar. These passes are used at 
 Neuberg in Styria, and are represented merely according to 
 their position on the rolls and in full size, while Figs. 66 
 and 67 show the individual passes in full size. The inverted 
 position of these passes, and the reasons for choosing the 
 same, have already been the subject of remark. A notable 
 point is the unusual height of the pitch line, as shown in 
 Fig. 65 ; this somewhat excessive height was, on account of 
 the inversion of the passes, considered necessary to prevent 
 the bar from curhng upwards. The rolls may be useful in 
 case of necessity, but are scarcely adapted to production on 
 a manufacturing scale. 
 
 The passes used at Seraing, in Belgium, for a peculiar 
 variety of angle iron, are represented in Fig. 83, on Plate 
 VIII., and the individual passes in full size, in Fig. 84. It 
 is very doubtful whether these passes are properly construct- 
 ed, and it seems improbable that the fifth pass can be cor- 
 rectly proportioned. The drawings were received directly 
 from the works. 
 
 All varieties of angle iron necessitate very good material, 
 on account of the difficulty of their manufacture and the 
 
TO 
 
 nature of their subsequent uses ; hence, the iron generally 
 used is that which has already been once or twice reheated 
 and rerolled. The piles for reroUing for this purpose, are 
 quite frequently built up of good scrap-iron. 
 
 § 28. The passes used for T iron resemble those used for 
 T rails. The latter are, however, in general more difficult to 
 roll than the former, at least in the forms in which the former 
 are applied in architectural and engineering purposes, and the 
 difficulty of rolling the T rails becomes comparatively greater 
 when different irons are used for the head and flange of the 
 T rails. Therefore fewer passes are usually required for 
 rolling T iron than for T rails. 
 
 T iron is required in much larger sizes than are the rails ; 
 which fact not only renders it difficult to proportion the 
 passes properly, but also makes the rolling very laborious. 
 The large size of the passes necessitates heavy rolls, espe- 
 cially bottom rolls, the diameter of which must often be 30 
 inches or more ; as such heavy rolls are very expensive, it is 
 advisable to reduce the side-thrust as much as possible by 
 the use of flatting passes. 
 
 Figs. 75 and 76, on Plate VII., represent respectively the 
 roughing and finishing for heavy T iron, as produced at 
 Piela, in Upper Silesia, while the passes are appended in full 
 size in Figs. 77 and 78. Their similarity to the passes for T 
 rails is very striking, while the presence of 3 or even 4 flat- 
 ting passes out of 10 passes in all, is equally worthy of 
 remark. The position of the first pass is that of a flatting 
 pass, and if this is considered as such (although this is sel- 
 dom the case with the first pass), we actually find 4 flatting 
 passes in use. 
 
 Figs. 79, 80, and 81, on Plate VIII., represent merely the 
 passes for a medium T iron ; these are constructed according 
 to Talbot's drawings, are in full size, and the pitch line is 
 shown. The drawings represent the two first passes in the 
 position of flatting passes, as was the case with the above 
 rolls, but the 6th pass is the only strictly flatting pass in the 
 rolls. There are 8 passes in all. 
 
 Double T or girder-iron (I beams), the top and bottom 
 
71 
 
 flanges of wliicli are similar in size and shape, is mucli easier 
 to roll than simple T iron, because the metal of the former is 
 equally divided on each side of the vertical centre line, and 
 hence the respective side-thrusts are equal. As no flatting 
 pass is necessary in rolling girder-iron, and as the beam is 
 merely turned half over (180°) at each pass, the requisite 
 passes are quite easily turned. It is, however, difficult to 
 roll beams so large as they are sometimes demanded ; the 
 rolls must therefore be very heavy, and a reversing train 
 should be used. 
 
 Fig. 82, on Plate VIII., represents in full size the passes 
 for an I beam 10 inches high, as they were drawn immedi- 
 ately from the turning tools ; these rolls are in use at Kes- 
 chitza. The remarks in § 21 upon the construction of piles 
 for I beams, as illustrated in Fig. 20, B, have special refer- 
 ence to these rolls, the first pass of which is a shaping pass. 
 
 It is unnecessary to adduce special examples of rolls for 
 girder-iron, etc., of smaller sections than the above, as the 
 corresponding passes are not only much simpler than those 
 required by the larger sections, but are also very similar to 
 those to be immediately described, which are required for 
 channel-iron and heavy I beams. 
 
 § 29. Passes for rolling channel-iron and very heavy I 
 beams, are quite similar to the flat presses described in §§ 12 
 and 13 and drawn m Figs. 16 and 17, on Plate II., and so far 
 as the roughing rolls are concerned, the respective series of 
 passes are exactly similar. Therefore Fig. 85, on Plate VIII., 
 represents merely that part of the finishing rolls which con- 
 tains the final passes for an 8-inch channel bar, while the 
 passes themselves are given in full size in Fig. 86. The latter 
 figure contains the section of the bar as it comes out from the 
 last roughing pass ; this, as well as the four finishing passes, 
 is arranged in the proper position with reference to the pitch- 
 line. The draw and spread in the respective passes are 
 shown distinctly in the drawing. 
 
 The section of heavy I beams is so simple, and the rolls 
 employed in their manufacture so similar to those used for 
 channel-iron, that it seems quite sufficient to draw, in Fig. 
 
 6 
 
Y2 
 
 87, the full-size passes required for rolling tlie former section. 
 As the figure shows, only 3 shaping and finishing passes are 
 employed instead of 4 in the case of channel-bar, although 
 the mass of metal is greater in the former than in the latter 
 section. The reason for employing more passes for the sim- 
 pler section lies in the fact that the side-thrust of the upper 
 and lower grooves is quite unequal. It is evident from the 
 respective drawings that the heavy girder beam, as well as the 
 channel bars, must be turned half over (180°) after each pass. 
 The passes of both the above are given according to Bieder- 
 mann's drawings, and are in use at a Styrian mill. 
 
 The series represented in Figs. 88, 89, and 90, on Plate IX. 
 (as used by Talbot), is somewhat unlike the above, and resem- 
 bles the series for great I beams, described at the end of § 27. 
 The first roughing passes are constructed with reference to 
 the form of finished section, in order that the work upon all 
 parts of the section may be as uniform as possible ; the qual- 
 ity of the iron is thereby improved, especially in the flanges, 
 and the strength of the beam increased. Each pass is rep- 
 resented in its proper position with reference to the pitch line, 
 and their unequal division above and below this hne is neces- 
 sitated by turning the bar half over after each pass. The 
 passes for I beams, represented in Fig. 82, on Plate VIII., 
 are placed with reference to the centre line of each pass, in 
 order that the draw of the respective passes may be clearly 
 shown. 
 
 The five roughing passes of Fig. 88, on Plate IX., are com- 
 mon both to the channel bar and heavy I beam, as are also 
 the first two finishing passes in Fig. 90. The third finishing 
 pass in Fig. 90 completes the section of the heavy beam, 
 while the passes drawn in Fig. 89, and marked respectively 
 3 and 4, are used to finish the channel bar, and the above 
 third pass is skipped ; thus the heavy beam demands one 
 pass less than the channel bar, as was shown above in Figs. 
 86 and 87, on Plate YIII. 
 
 The European form of the chairs for T rails with a fish 
 joint is somewhat similar to that of channel-iron. These 
 chairs {vide § 21) are rolled in long bars, which are afterwards 
 
Y3 
 
 sawed up to the requisite lengths. The ordinary section of 
 the chairs is that of A in Fig. 29, although 
 it is a frequent practice on Continental 
 roads to use plates of the section B on 
 
 heavy curves where it may seem neces- 
 sary to regulate the angle of the rail to 
 the tie. These forms are rather harder 
 to roll than the ordinary channel iron, and therefore the 
 requisite passes are shown in Fig. 91, on Plate IX., in their 
 proper position in the roUs, while Fig. 92 represents them 
 in full size. 
 
 The difficulty in rolling such forms increases with the 
 height of the sides or lips, and it is therefore necessary to 
 leave ridges on the back of the bar {vide § 22), or on the side 
 opposite to that on which the lips are to be rolled. In order 
 to make this method perfectly clear. Fig. 93, on Plate IX., 
 represents the finishing and one roughing pass for the 
 manufacture of channel bar with a deep channel ; these 
 passes are shown in their proper position on the finishing 
 rolls, and are appended in full size in Fig. 94. The drawings 
 show that even the first shaping pass is turned with reference 
 to the formation of the ridges, which latter appear more and 
 more distinctly at each subsequent pass up to the last, during 
 its passage through which they are forcibly pressed down 
 into the sides of the channel bar. It is apparent that the 
 bar, as it leaves the last roughing pass, must be turned quar- 
 ter over (90") before it enters the first shaping pass. 
 
 § 30. Felly iron is an iron of half round form on one side, 
 and of wedge shape on the other, and demands attention on 
 account of its somewhat typical shape, although it is almost 
 exclusively used in the production of wrought-iron wheels. 
 It forms a support to the outer felly or felly proper over 
 which the tyre is shrunk, and is welded between the spokes of 
 the wheel, both to the outer felly and to the spokes. If the 
 ■wedge-shaped side is not too sharply pointed, it is easy to 
 develop the shape without the aid of a flatting pass, and 
 quite simply in the ordinary way, in which the bar is turned 
 half over at each pass. A series of such passes is repre- 
 
74 
 
 sented in Fig. 95, on Plate IX., and Fig. 96 sliows the indi- 
 vidual passes in full size witli reference to tlieir draw and 
 irrespective of tlieir position in the rolls, which is distinctly 
 given in Fig. 95. 
 
 If, on the contrary, the wedge-shaped side is very sharp 
 and thin, it is necessary to introduce a flatting pass, as in 
 Fig. 97, which represents in i^ full size, a set of rolls of a 
 Silesian mill ; the respective passes are given in full size in 
 Fig. 98. In this case the flatting pass is placed in the mid- 
 dle, and is only to be recognized by its position with refer- 
 ence to the rest, as it could not be recognized as such if con- 
 sidered singly. 
 
 It is easy to construct rolls for fish plates, as the form of 
 the latter is usually quite simple. Fig. 99, on Plate X., rep- 
 resents a pair of rolls for a very simple variety, while the full- 
 size passes are appended in Fig. 100. The shape of the 
 former of the second pass is rendered somewhat unusual by 
 the necessity of preventing fins from forming at the sharp 
 corners shown in the third pass, in which latter, however, the 
 draw and consequent tendency to form fins is very much less 
 than in the former pass. The sharp corners of a former are 
 avoided in the finishing pass, by giving this latter the form 
 of an open pass, but placing this open pass in the position 
 of a closed one and letting the former work between two col- 
 lars, as in the case of a closed pass. This union of the open 
 with the closed pass has been already noticed several times, 
 and is often exceedingly useful. The immediate outlines of 
 such a pass are those of an open one (i. e., there is a groove 
 in the former, so that in reality the former acts the part of 
 the top roll of any set with open passes), but considered in 
 its general relations and position, the pass must be denomi- 
 nated a closed one. 
 
 § 31. As an example of a very useful class of passes, it will 
 be well to consider those which are used in rolling rods of 
 pecuKar form, intended to be cut up into spikes. In this 
 country we use the spike machine almost exclusively, but it is 
 quite probable that we might find passes of a form similar to 
 these very useful for rolling a variety of other shapes which 
 
75 
 
 require periodical projections. The form of tlie subjoined 
 passes for spike rod is such that the notches correspond to 
 the heads of the single spikes. In order to fill out the notch 
 properly with the metal which is forced up into it, a very 
 heavy draw is required, and if the pass is narrow, fins are 
 very likely to be formed. In § 22, however, a method of 
 avoiding the latter was described, which is in use at Mayr's 
 mill near Leoben. 
 
 At Eeschitza the rods are finished in another way. Fig. 
 101, on Plate X., gives in | full size a front elevation and a 
 section of the rolls used in this mill. The circumference of 
 the rolls is equal to four times the length of a single spike, 
 and at every revolution the corresponding four notches form 
 the head of a single spike. The rod leaves the roughing 
 rolls with a rectangular section, | an inch thick by of an 
 inch wide, which dimensions remain unchanged in the head 
 of the spike. 
 
 In the finishing rolls five similar passes, a, a, and a very 
 wide pass, b, are used. Only one of the former (a) is used at 
 a time, but five are on the rolls, because the corners of the 
 holes of the dies for the spike head are soon worn out of 
 shape even though the pass is an open one. On account of 
 the open form of the pass, the fins formed will be in the mid- 
 dle of the side, and are made very blunt by rounding off the 
 comers a good deal. The drawing represents three passes 
 with a width of | an inch, which is also the thickness of the 
 rod as it reaches them ; they might be wider in order to 
 facilitate the introduction of the rod, but the extremely 
 rounded form of the collars renders such extra width unne- 
 cessary. In order to remove the fins from the shaped rod, 
 it is rolled through the wide flat pass, h, the height of which 
 is equal to the thickness of the spike rod, i. e., | an inch. 
 
 Fig. 102 represents in ^ a single spike as cut off of the 
 rolled rod, while B represents the form into which it is 
 brought by hand, and in which it is sold. The spikes might 
 be more readily and very quickly finished by a machine con- 
 structed for the purpose. 
 
 In order to show how the same form may be attained by 
 very different means, a third method is subjoined, which was 
 
76 
 
 in use at Fridau's mill, near Leoben. Fig. 103, on Plate X., 
 gives the front size of the rods in \ full size, and it will be 
 observed that the section is dotted in order to save space. 
 The rolls contaia 3 first and 3 second shaping passes which 
 are turned as closed passes, and the differences between 
 them are very small, as thej are not intended to be used 
 successively upon the same rod ; so many passes are neces- 
 sary to ascertain those which shall be used, according to the 
 heat and quality of the metal in the rod, in order to obtaia 
 the best results. Only one, or at most two, of each 
 set of shaping passes is used, while the height of the passes 
 is often slightly changed by raising or lowering the upper 
 roll. 
 
 The roUs contain only one intermittent pass, if it may be 
 called pass ; this one, however, is 15 inches long, and forms 
 the end of the body of the roll. Only a small part of the 
 whole width is used at a time, viz., so much as the rod takes 
 up ; the latter is kept in position by an accurate guide fitted 
 to and projecting between the rolls (similar to that drawn in 
 Fig. 34, on Plate III., and described ia § 13). As one part 
 becomes worn, the guide may be moved further along the 
 rod, thus rendering the pass serviceable for a much longer 
 time. This peculiar form has the additional advantage that 
 the notches may be quickly and easily formed in a planing 
 machine, whereas the notches in a narrow pass must be 
 finished with great difficulty by hand. 
 
 After the head has been pressed up in this intermittent 
 pass, or intermittent portion of the roll, it is turned quarter 
 round, i. e., upon its side, and passed through a pair of small 
 polishing-roUs, in which any fins are rolled down, and the 
 with of the bar between the heads is reduced by the amount 
 to which it spread in the intermittent pass. 
 
 Fig. 104 represents in A the rough spike as cut from the 
 rod, and in 5 the spike as finished by hand. A very simple 
 method of avoiding fins, though one which is little known 
 and seldom used, is shown in. the shaping passes of these 
 rolls (Fig. 103). The sides of the collars are flared out in- 
 stead of being left straight, and the bar is turned haK over 
 at each pass, thus bringing the narrow bottom of the bar 
 
T7 
 
 into the wide top of the next pass, which top is so much 
 wider than the bar that scarcely any draw will force the 
 metal out into fins. This method is very efficacious, and de- 
 serves universal application in rolling flat and square bar or 
 hoop iron, as it leaves the comers perfectly free from fins. 
 
 § 32. The method pursued in turning passes for a variety 
 of small shapes is illustrated by Figs. 105, 106 and 107, on 
 Plate X., which represent four different series of passes for 
 sash iron, constructed according to drawings from Seraing. 
 The passes are drawn m full size and in their proper posi- 
 tion on the rolls, but the rolls themselves are not represented 
 further than is necessary to render their completion easy. 
 The radii (half diameter) of the rolls are, therefore, given in 
 figures on the end collars. 
 
 In order to avoid the formation of fins, all the sharp 
 corners and edges are rounded off, and the passes are so 
 placed that the bar must be turned each time half over 
 (180°). They are also so proportioned as to allow for a 
 small spread, so that the bar may enter the pass easily. It 
 would also be well to flare the side of the passes somewhat, 
 as in Fig. 103, particularly if the stock rolled was red short, 
 or of low quality. 
 
 As the roughing rolls for these shapes contain only Gothic 
 and flat passes, it is merely necessary to refer to previous 
 examples of similar rolls. 
 
 In rolling such small shapes, it is very necessary to use 
 guides which are closely fitted to the rolls and to the pass, 
 in order to prevent the destruction of or cutting off the 
 corners of the bar as the grooves take hold. The diameter 
 of the top roll in the grooves is throughout comparatively 
 great, as the drawings show, in order that the bar may have 
 a strong downward tendency, while accurate guards are used 
 and water is plentifully appHed. These precautions are ne- 
 cessary to prevent the bar from curling round the roll. 
 
 § 33. Hitherto our attention has been confined to the more 
 usual varieties of passes, but the definition of the term pass, 
 
78 
 
 given in § 2, would require that many other and quite differ- 
 ent arrangements for rolling iron should be discussed. To 
 treat the subject fully, however, would be almost impossible 
 in a work of this character, especially as the number of draw- 
 ings would be largely increased. It wHl, therefore, be suffi- 
 cient to describe the principal varieties of machines for 
 rolling or sHtting iron, which cannot be strictly classed as 
 rolls with passes. 
 
 The 81itting Mill is a very important and verj- generally 
 used machine for sHtting or dividing, at a single pass, flat 
 bars or sheet-iron into a number of rods, as nail rods or sheet 
 for welded boiler tubes. Generally 10 or more of the smaller 
 rods are slit off at a time, with the length and thickness of 
 the origmal bar or sheet, but with a width which is regulated 
 by the distance between the cutting discs. These discs 
 operate exactly as circular shears, but many of them are 
 united, as it were, in one piece, the middle discs cutting on 
 both edges. The discs are forged singly, either wholly of 
 cast-steel or of iron, to which steel is welded, to form the 
 edge. Between two of these discs is placed another of the 
 same or any desired thickness, but less diameter, and the 
 whole are fixed concentrically upon an iron spindle, which is 
 provided with the necessary neck and pods. These discs 
 and short cyHnders may be fastened by being driven up by 
 screws, or a short, heavy cotter or wedge against a fixed col- 
 lar at the other end of the spindle. When, however, the 
 width of the desired sKt sheet is considerable, it is customary 
 to use cast-iron rolls, upon which collars are cast of the ne- 
 cessary diameter, and at such distances that when faced on 
 the circumference with rings of steel, they will allow the steel 
 rings on the collars of the top and bottom rolls to work close 
 up to each other, thus making the slit. The collars are ordi- 
 narily faced with semicircular rings, fitted on to a turned 
 face on the collar, and fastened with counter-sunk bolts. 
 These mills must be very accurately constructed, and require 
 a good deal of attention in use, but do their work very 
 quickly. They are generally placed at the finishing end 
 of the train, that the sheet or bar may be slit immediately 
 on leaving the poHshing rolls or the last pass. 
 
Y9 
 
 The so-called "collar rolls" are used to widen out any 
 individual part of a flat bar wliicli is intended to be bored 
 for a bolt. These bars are generally those used for forming 
 the links of solid chain suspension bridges, and often require 
 to be widened at each end as well as in the middle. In 
 order to be able to roll at will such wider parts, the collar 
 rolls are generally constructed of a wrought-iron spindle, 
 upon which, at the required intervals, wrought-iron collars 
 are shrunk. These collars may be readily moved by being 
 heated by a thick iron ring, laid around them in segments at 
 a white heat ; the consequent expansion loosens their hold on 
 the spindle, and they may be moved as desired. The spindle 
 may, of course, be made of cast-iron, and where a demand 
 for special forms exists, the necessary collars may be cast 
 solid with the spindle. In the practice of Howard, Eaven- 
 hill & Co., of Rotherhithe, the spindles are wrought-iron, 7 to 
 8 feet long, and turned to a diameter of 6 to 7 inches. They 
 are, of course, furnished with necks and pods, which are in- 
 cluded in the above length. The wrought-iron collars are 
 shrunk on the cylindrical body of the spindle, and are not 
 rectangular, but have a rounded or convex surface, in order 
 that there may be no abrupt depression in the bar. In roll- 
 ing, the bar is passed three or four times through the rolls, 
 while at each passage the top roll is lowered. The collar roll 
 housings are placed at the end of the train in which the bars 
 are rolled, in order that the latter may be finished at a single 
 heat. If arranged to reverse, they are placed at right angles 
 to the end of the train, and driven from the respective rolls 
 by an arrangement of cone pinions, constructed to throw in 
 and out of gear at will, and thus reverse as desired. 
 
 The term "end roUs," or "overhang rolls," may be applied 
 to those rolls which are merely spindles between the necks, 
 and whose body is merely a short continuation of the spindle 
 beyond one neck, while the other end of the spindle is con- 
 nected with the engine by means of gearing. The short 
 body or head of the roll contains only one pass, that is, the 
 former is turned upon the upper head and the groove into the 
 lower one, or the groove on the stationary and the former 
 
80 
 
 upon the movable, when the spindles are arranged as in an 
 ordinary tyre mill. 
 
 The reason why such an arrangement is necessary is be- 
 cause these end rolls are required for rolling tyres, etc., the 
 form of which objects is a closed ring, which must be placed 
 in position and removed quite easily ; one side of the pass 
 must, therefore, be left entirely open, or at least be easily 
 opened by moving back the movable spindle. The draw of 
 these passes or rolls is given by continually moving the rolls 
 together, or the movable against the stationary head, while 
 the tyre, or other object, is being rolled. As a very con- 
 siderable force is required to move the rolls together, the 
 movement cannot be accomplished by hand, and it is neces- 
 sary to employ an arrangement with friction pulleys, by 
 which the rolls may themselves work the shifting screw or 
 screws, or a small steam-engine, or better, a hydraulic press, 
 to raise the journals of the bottom roll, or press the movable 
 one up. 
 
 As the tyre is rolled out its diameter becomes greater, and 
 the cross-sectional area of the metal becomes proportionately 
 smaller, as well as the difference between the external and 
 internal diameters. The speed of the rolls must, however, 
 vary between themselves and become more nearly equal as 
 the difference between the diameters grows less, in order to 
 avoid an irregular draw and even cutting away of the metal ; 
 hence the end rolls are often provided with from 2 to 4 
 passes, the relations of the diameters of the grooves of which 
 are different. The diameters of the grooves are so arranged 
 that they may at first draw the outside or tread the most, and 
 that each succeeding pass may draw the same more nearly 
 equally with the inner surface. Another arrangement con- 
 sists in using two separate pairs of end rolls driven by the 
 same motor, and with similar relations of draw. The latter 
 method has the advantage of finishing and centring the tyre 
 at the same time. 
 
 The axes of the spindles may be either horizontal or verti- 
 cal. The horizontal spindles are in use at Blaenavon and 
 Eotherham, in England, and Seraing, in Belgium, as well as 
 
81 
 
 in several Frencli mills, and at Stefanau, in Moravia. In the 
 vertical arrangement, one roll is fixed in position and merely 
 rotated, while the other, in addition to its rotary movement, 
 is also moved toward the former ; the latter roll is generally 
 the inside one, i. e., that which forms the inner surface of the 
 tyre. This arrangement, with two passes on the rolls, is in 
 use at Monkbridge, in England, while a similar machine at 
 Bochum has three or four passes on its rolls. In changing 
 from one pass to another, the rolls must of course be opened 
 wide and the tyre lifted into position. 
 
 The "Universal Mill" has been already mentioned (§12) as 
 the best means of saving a great variety of simple flat passes 
 and consequent stock of rolls. It has been described in 
 many technical journals, but especially well by the inventor 
 himself in "Dingier s Polytechnisches Journal" vol. 164, pp. 401- 
 403. The inventor of this arrangement of rolls is Daelen, of 
 the Horde mill {vide § 12). Mr. Wagner, of Maria Zell, has 
 made some useful improvements upon the original, and built 
 quite a large mill of this kind. The mill itself consists of two 
 horizontal rolls which can be closed and opened, and imme- 
 diately in front of them two vertical rolls to act upon the 
 edges of the bar. The various improvements consist chiefly 
 in the mode of applying and regulating the gearing which 
 moves the respective rolls. 
 
 This mill is not well adapted for rolling very thin iron, 
 because such bars or rods become quite cold in the last few 
 passes, and are thus proportionately more spread (§ 7) side- 
 wise ; which spread the vertical rolls cannot reduce by any 
 means so easily as they could at first, when the heat was 
 high and the spread inconsiderable. If the bar were thin 
 and the spread considerable, it would be almost impossible 
 to prevent it from being bent (in the direction of its width) 
 between the vertical rolls. 
 
 The rolls used for tubes with thin walls and large diame- 
 ter, have the peculiarity that the pass is divided upon four 
 rolls, which are of segmental form, and the surfaces of which 
 form 90^ of a circle, corresponding to that of the outside of 
 the tube, while a mandrel carried on a long stem forms the 
 
82 
 
 inner circular surface. If the draw were heavy and the walls 
 of the tube thin, it would be impossible for two ordinary- 
 geared rolls fthe passes of which have a varying surface 
 speed) to bring the tube over the mandrel ; this is, however, 
 easily accomplished by the four rolls geared so together that 
 their surface speed is quite equal. Each of the four seg- 
 mental rolls contains, as above, one quarter of the required 
 circle, two being placed horizontally and two vertically. 
 
 § 34. The manner in which rolls are turned up in the lathe 
 may be succinctly described as follows : 
 
 The requisite form must, of course, be laid out on paper in 
 detail and in full size. Templates are then formed in sheet 
 iron, or brass, exactly accordmg to the drawing, and the ne- 
 cessary tools are also, for the most part, shaped to corre- 
 spond with the various parts of the drawing. New tools, 
 however, are only needed for those parts which are not 
 usual, such as the bottoms of grooves and faces of formers ; 
 for rounding off the corners of the body-fillet, etc., 
 and for forming many other parts, no special tools are 
 necessary, as the forms of such parts are those which often 
 occur in aU rolls. 
 
 The templates may be prepared either for each pass singly, 
 
 or one general tem- 
 plate may be cut 
 out for the whole 
 body of the roll, 
 i. e., all grooves, 
 formers, collars, 
 or body fillets 
 which may occur 
 thereon. In Fig. 
 30, A and B re- 
 present templates 
 for single passes, 
 while C includes the whole body of the roll. Templates are 
 used to give the exact form of a pass, as well as its proper 
 
 Fig. 30. 
 
83 
 
 position on the rolls ; therefore, a template of the form B is 
 obviously insufficient, and is accordingly used merely for 
 convenience, so as to avoid frequent handling of such a large 
 plate as C. The form of template shown in A is used to 
 determine the exact position of any given groove, the part a 
 being pressed against the outside of the end-fillet, and the 
 parts c c lying on the body-fillets or collars, as the case may 
 be ; as this form necessitates a number of unwieldy templates, 
 the form C should be used in preference. One template 
 suffices for both rolls, if the passes used are open and equally 
 divided at the pitch hue, since the small differences of the 
 diameters of the respective grooves in each roll are without 
 influence on their form. It is, on the contrary, necessary to 
 employ two templates, or to utilize both sides of the same 
 one, as in D, Fig. 30, if the passes are of the closed form, 
 as these are scarcely ever divided equally at the pitch line. 
 
 The roU, as it comes from the sand, is placed between the 
 centres of a lathe, and the sinking head or riser cut off; it 
 must then be accurately centred and the necks turned up 
 roughly, to ascertain whether or not any fault or flaw exists, 
 which would be sufficient to condemn the roll ; if there is 
 none, the necks are finished. It is most convenient to rough 
 out and finish the necks on a lathe with feed, and then to 
 transfer the roll into the regular steady rest to rough out and 
 finish the passes. In order to turn up the passes, it is ne- 
 cessary to support the roll on its necks, and so firmly that it 
 will not be sensibly jarred by the tool, which presses with 
 considerable force against it. The necks are, therefore, sup- 
 ported in " steady rests," constructed somewhat similarly to 
 the housings in the train, that the roll may be supported 
 against thrust in any direction. 
 
 The steady rest shown in Fig. 108, on Plate VIT., is well 
 adapted for turning small rolls, on account of the firmness 
 with which the roll can be held in position ; the form is anti- 
 quated, however, and it is introduced as an illustrative type 
 of its class rather than as an example for practice. The base 
 a rests on the bed of the lathe, and is fixed in position by 
 the cross-piece c, which is drawn firmly up against the under 
 
84 
 
 side of the bed by means of tbe screw 6. TLe top is left 
 open, and is closed when necessary by the bar d, which is 
 held down, and in place, by the screws e e. The inner sides 
 of the uprights are quite smooth, and the edges rectangular, 
 and the side chocks are constructed to overlap them some- 
 what on the inner side. The top and bottom chocks h h are 
 let into the bars i i, which are here represented as wood, but 
 which it is better to make hollow and of iron. If the roll is 
 small, it is well to leave it stiU between the lathe centres, 
 although supported in the steady rest, in order that the roU 
 may be immovable endwise ; if large, however, it is driven 
 either by a sleeve, or by a casting resembling haK a sleeve, 
 which is bolted on the face plate, while any movement end- 
 wise is prevented by the portion of the end-fillets against the 
 chocks. The screws A; ^ at the sides, m at the top, and the 
 slender wedge I at the bottom, set up the chocks to their 
 exact position. 
 
 One of the simplest and most practical forms of " steady 
 rest " or " housing " for aU kinds of rolls is that shown in 
 Fig. 31. Here there is an U-formed standard, from the bottom 
 of which a strong flange projects on each side ; the lower 
 face (h) of this flange is planed smooth and rests on the bed 
 of the lathe, being usually held in position with bolts, so 
 arranged in various ways as to be easily shifted. At the 
 junction of the sides of the standards there is a groove, in 
 which the piece a moves up and down, being held in position 
 by the bolt c and block of wood e, and being moved by the 
 set screw d. This piece (a) supports the neck of the roll. At 
 the top of each of the sides of the standards there is an 
 overhanging lug, the inside of which projects downwards 
 lower than the outside. Through this lug the bolt g projects 
 and works in a slot in the piece h, which it can firmly hold 
 in any required position, when the piece h has been moved 
 into that position by means of the set screw /. All the sur- 
 faces of b and the lug which come in contact with each 
 other are, of couse, carefully finished. These pieces b pro- 
 ject downwards against the neck of the roll, thereby prevent- 
 ing it from springing upwards, though not opposing its rotary 
 motion. The pieces b are called "jaws," and in order 
 
86 
 
 tliat they may be in the same vertical plane as the rest a , 
 the upper part of the latter projects so as to come mider 
 them. If the axis of the roll, on being placed on the steady 
 rest, does not coincide with a line drawn between the cen- 
 tres of the lathe, the rest a may be raised till the roll is in 
 the proper position. 
 
 Fig. 31. 
 
 Tools of hard-chilled iron are sometimes useful for the 
 first rough cuts, to remove the hard skin ; but these are sel- 
 dom used, and it is better to make all of good cast-steel. 
 For small work, both ends of the tool are arranged to cut as 
 
86 
 
 sSdwn in Fig. 32, A ; for heavy work, however, sucli a form 
 is not advisable, as the tool must in one form of rest be 
 blocked firmly into, and placed and forced against, the roll 
 by a powerful screw, acting on the flat end of the tool, and 
 in another form of rest be shaped like an ordinary lathe tool. 
 
 It is advisable to weld a plate of cast-steel (as the cutting 
 edge) upon an iron tool, as the exclusive use of steel, ex- 
 cept in the case of smaller tools, would be quite expensive. 
 
 32. It is sometimes advisable that 
 
 ^ W cuttmg piece should be 
 
 merely inserted in the tool, and 
 not welded upon it. For ex- 
 
 ^ ample, as the passes tor light 
 
 round iron must be turned ex- 
 actly circular (§16 and § 18), 
 
 , „ it is best to prepare in the 
 
 LB 
 
 1 ™ lathe a cyhnder of steel of the 
 jj. exact diameter of the required 
 ™ circle; then cut the same up 
 into small cylinders ; plane the ends and harden the pieces. 
 These hardened cylinders are let in for about half their 
 length into the end of the tool which has been formed for 
 their reception, as shown in B, Fig. 32. Those oval passes, 
 which partially consist of two similar arcs, may also be best 
 turned in this way, by merely cutting in to the necessary 
 depth. The cylinder may be turned round as its edge be- 
 comes worn, and the edges of the other end may be subse- 
 quently used. When the edges of both ends are worn out, 
 they merely require grinding to restore them to the proper 
 shape. 
 
 The large roughing passes are usually cast in the roU quite 
 near the finished size and form, and it might probably be 
 weU to leave the largest grooves untouched, as they do not 
 require any very great accuracy, would soon be worn smooth, 
 and the hard skin would greatly improve their wear. 
 
 If passes of large size must be turned wholly out of the 
 body of the roll, it is well to use at first a simple tool, which 
 may be easily made and repaired, and to finish with the 
 
special tools, whicli are prepared with consict&^hleff^^^^ 
 culty. 
 
 In turning closed passes, the formers of which must fit 
 with as little play as possible into their respective grooves, 
 it is well to use a template for the whole body of the roll. 
 When the rolls are finished it is well to place one over the 
 other, the latter remaining in the steady rest, just as it was 
 turned. To facilitate this, the sides of the rests may be 
 made high enough to take in the roll, and carry it as in an 
 ordinary housing, or any special carrying arrangement may 
 be placed over the steady rest. This comparison of the rolls 
 facilitates the discovery of any mistakes in turning, and gives 
 an opportunity to remedy faults, or improve the form, as the 
 case may be. We have already seen that the sides of the 
 grooves of closed passes are not exactly perpendicular, but 
 flare outwards to the extent of to of an inch or more, 
 that all sharp corners and angles of the groove and former 
 must be avoided as far as possible, and that it frequently 
 happens that it is necessary to alter a set of rolls according 
 to the results of their first trial in the train, as these are 
 sometimes unsatisfactory. 
 
 7 
 
89 
 
 TECHNICAL TEEMS 
 
 KOLL-TURNINa, 
 
 ENGLISH, GEEMAN, AND FEENCH. 
 
 Apron, vicl. Bearing bar. 
 Bearing bar. 
 
 Body-fillet. 
 Box or Sleeve. 
 
 Box groove. 
 
 Brasses, md. Journal-box. 
 Breaking box. 
 Breaking shaft, vid. Spin- 
 dle. 
 Chilled rolls. 
 
 Chock. 
 
 Clutch. 
 Collar. 
 
 Coupling crab, vid. clutch. 
 Diamond pass. 
 Draw or Draught. 
 Edge pass. 
 End-fillets, 
 
 Feed plate, vid. Bearing 
 
 bar. 
 Fillet. 
 
 Finishing pass. 
 Finishing rolls. 
 
 Fin. 
 
 Flat pass. 
 Flatting pass. 
 
 Walzenbank, 
 
 Kuppelimgii* JSIt^q, ; of 
 simply Mufi'd • *.* .* • J 
 
 Flach Kalibej: (one va^ 
 riety of). ;/,* . , 
 
 BrechkapseF, Bjec^iJck. 
 
 Hartwalzen. 
 Lager, Zapfenlager. 
 
 Keilmuff, Kuppelungs- 
 
 scheibeu. 
 Walzenring-ring, Eippe. 
 
 Quadi'at kaliber. 
 Abnahmsverhaltniss. 
 Stauch kaliber. 
 Endrtng. 
 
 Walzenring. 
 
 VoUend kaliber. 
 Vollend walzen, or 
 Schlicht walzen. 
 Bart. 
 
 Flach kaliber. 
 Breitungs kaliber. 
 
 Tablier, or before the rolls, 
 Plague des guides. 
 
 Tablier, or hehiiul the rolls, 
 Plague des gardes. 
 
 Cordon. 
 
 Mjinchon d' accouple- 
 
 ^ft^iffc- *.*• 
 •G^pijelurtf rsigt^w^ulaire. 
 
 • • • 1 * .* * * ' 
 
 Cylinders* ct)nl6? 'eij ecL- 
 
 quilles. * * * * 1 
 
 Above all necks, Chapeau 
 
 de cage. 
 Between or below the 
 
 necks, Tourillets. 
 fichappement, manchon 
 
 a griffes. 
 Collet. 
 
 Cannelure carries. 
 Tirage (Pression). 
 
 Cordon. 
 
 Cordon. (Cordon des 
 
 cylinders femelles. ) 
 Cannelure finisseuse. 
 
 Cylinders finisseures. 
 Ger9ure. 
 Cannelure plate. 
 Cannelure de champ. 
 
90 
 
 EOLL-TURNING— Continued. 
 
 Former. 
 
 Gothic pass. 
 Groove. 
 
 Groove and Former. 
 
 Guards. 
 Guides. 
 Guide rolls. 
 Housing (train). 
 
 Housing (lathe), vid. 
 
 Steady rest. 
 Journal-box. 
 Merchant rolls. 
 
 Mill bars. 
 Neck. 
 
 Oval pass. 
 
 Pass. , *. ; •, '. 
 Pitch* linV.;* * '.* • * 
 •.tlUe.rqJla.:' • •** 
 jPoS. ' * , . ; 
 
 f*uddle rolls. • * '• ' *. ; 
 Eedy.Qtiipn/^yifZ^^DmV. • • 
 E5»]igJSiig *p*f>sa* • * * 
 
 Roughing rolls. 
 
 Shoe (train). 
 Sleeve, vid. Box. 
 Spindle. 
 
 Standard, vid. Housings. 
 Step rolls. 
 
 Steady rest (lathe). 
 Table, vid. Bearing bar. 
 Train. 
 Tyre rolls. 
 
 GEBMAN. 
 
 Patrize, Kahberring. 
 
 Spitzbogenkaliber. 
 Einschnitt, Matrize. 
 Matrize and Patrize, 
 
 Abstreifmeisel. 
 EinlSsse. 
 Feineisenwalzen. 
 Walzen geruststander,or 
 simply Stander. 
 
 Einsetzlager. 
 
 Grobeisenwalzen, or 
 Feineisenwalzen, or 
 Fa9oneisenwalzen. 
 
 Rohscheinen. , 
 
 Lagerzapfen ^qv' Jjauf- 
 zapf^n. *. * * 
 
 .OyalTiifljbef^ ;' *. ' 
 
 JJ^aliVer. 
 
 MitteUinia*. 
 
 BlftchVa^fefi. 
 ^ Ku^ ji^u\igsf;apfen. 
 *J^UP{)enwalzen. 
 
 Vorkaliber, or Streck- 
 
 kaliber. 
 Vorwalzen, or Streck- 
 
 walzen. 
 Sohlplatte. 
 
 Zwischenwelle. 
 
 Stufenwalzen, Staffel- 
 
 walzen. 
 Lunette. 
 
 Walzenlinie. 
 Tyreswalzen. 
 
 Rondelle, or Cordon des 
 
 cylinders males. 
 Cannelure ogive. 
 Cannelure. 
 
 Cannelure and Rondelle, 
 or Cannelure des cylin- 
 ders femelles. Cordon 
 des cylinders males. 
 
 Gards, Racloirs. 
 
 Guides or Gides. 
 
 Gid rolls. 
 
 Cage. 
 
 Coussinet. 
 
 Cylinders marchands. 
 
 Fer (5bauch^. 
 Tourillon. 
 
 Cannelure elliptique, or 
 
 Cannelure plateuse. 
 Cannelure. 
 
 Cylinders k t61e. 
 Bout or Trifle. 
 Cylinder ^baucheurs. 
 
 Cannelure degroisisseuse 
 
 Cylinders degroisisseurs. 
 
 Lit. 
 
 Allonge. 
 
 Poup^e a lunette. 
 Jeu. 
 
 Cylinders a bandages de 
 chemin de fer. 
 
T -A. B L E 
 
 SHOWING IN ENGLISH INCHES THE EQUIVALENTS OF AUSTEIAN 
 INCHES AND PARTS OF INCHES. 
 
 AUSTKIAN. 
 
 English. 
 
 Austrian. 
 
 English. 
 
 Austrian. 
 
 English. 
 
 1' = 
 
 2' = 
 3' = 
 
 1' 0.445" 
 2' 0.891" 
 3' 1.336" 
 
 4' = 
 5' = 
 6' = 
 
 4' 1.782" 
 5' 2.227" 
 6' 2.673" 
 
 7' = 
 
 8' = 
 9' = 
 
 7' 3.118" 
 8' 3.564" 
 9' 4.010" 
 
 Austrian Inches 
 IN twelfths. 
 
 1" 
 
 ^1-" 
 
 T2 
 _2_" 
 1 2 
 3 // 
 
 Ji-I' 
 
 T2 
 5 // 
 
 T2 
 
 7 '/ 
 
 8 // 
 T2 
 
 T2 
 10// 
 T2 
 11// 
 
 T3" 
 
 1 // 
 
 _2_// 
 1 2 
 _3_// 
 T2 
 
 3_// 
 
 T2 
 
 J5_// 
 
 12 
 
 T%" 
 
 A" 
 ^§_// 
 
 T2 
 
 J_// 
 
 T2 
 10// 
 
 2" 
 
 T2 
 
 2 // 
 T2" 
 
 3 // 
 
 English Inches 
 in decimals. 
 
 0.0864275' 
 
 .172" 
 
 .259" 
 
 . 345" 
 
 .432" 
 
 .518" 
 
 .604" 
 
 .691" 
 
 .777" 
 
 .864" 
 
 .950" 
 1.037" 
 
 .123" 
 
 .209" 
 
 .296" 
 
 .382" 
 
 .469" 
 
 .555" 
 
 .642" 
 
 .728" 
 
 .814" 
 
 .901" 
 
 .987" 
 2 . 074" 
 
 .160" 
 
 .247" 
 
 .333" 
 
 Austrian Inches 
 in twelfths. 
 
 2" 
 
 4 // 
 
 T2" 
 _5 // 
 
 6 // 
 
 7 // 
 
 // 
 
 3" 
 
 9 // 
 TY 
 IJL" 
 
 1 2 
 11// 
 
 4" 
 
 _i_/' 
 
 1 2 
 _2 // 
 1 2 
 ^3.// 
 T2 
 4 // 
 1 2 
 6 // 
 
 TIT 
 
 __6 // 
 T"5- 
 7^// 
 
 1 2 
 
 .9." 
 T2 
 10// 
 1 2 
 11// 
 1 2 
 
 ■^" 
 T2 
 
 ^" 
 
 3 // 
 TT 
 
 4 // 
 
 T2- 
 
 6 // 
 T2" 
 6// 
 
 English Inches 
 in decimals. 
 
 2.419" 
 . 506" 
 . 592" 
 .679" 
 . 765" 
 .852" 
 . 938" 
 
 3.024" 
 .111" 
 .197" 
 .284" 
 .370" 
 .457" 
 . 543" 
 . 629" 
 . 716" 
 .802" 
 . 889" 
 .975" 
 
 4.062" 
 . 148" 
 .234" 
 .321" 
 .407" 
 .494" 
 .580" 
 .667" 
 
92 
 
 TABLE OF AUSTRIAN AND ENGLISH mCimS- Continued. 
 
 AUSTEIAN. 
 
 English. 
 
 Austrian. 
 
 English. 
 
 Austkian. 
 
 English. 
 
 1' = 
 
 2' = 
 3' = 
 
 1' 0.445" 
 2' 0.891" 
 3' 1.336" 
 
 4' = 
 5' = 
 6' = 
 
 4' 1.782" 
 5' 2.227" 
 6' 2.673" 
 
 7' = 
 8' = 
 9' = 
 
 7' 3.118" 
 8' 3.564" 
 9' 4.010" 
 
 ArsTEiAN Inches 
 
 English Inches 
 
 Austkian Inches 
 
 - 
 
 English Inches 
 
 IN twelfths. 
 
 IN DECIMALS. 
 
 IN TWELFTHS. 
 
 in decimals. 
 
 All 7 // 
 ^ T2 
 8 '/ 
 
 T% 
 
 4: . ( Do 
 
 7/' 6 
 
 7 77S" 
 
 , OOU 
 
 
 . 864 
 
 TTT 
 1 0 
 
 Ta" 
 
 
 8 
 1 2 
 
 . t/OX 
 
 019" 
 
 0 
 
 1 -2 
 
 8 0^7" 
 
 1 1 
 
 0 
 
 , yj 1 iJ 
 
 1 0 
 12 
 
 
 . XoO 
 
 1 1 
 
 \2 
 
 91 0 
 
 1 
 
 m 
 
 979 
 
 8" 
 
 9Q7 
 
 2 
 
 ■nr 
 
 . ooo 
 
 
 . 000 
 
 
 AAA 
 . 44:4: 
 
 A 
 
 /I 
 
 .4:Dy 
 
 
 . UOX 
 
 3 
 
 T2- 
 
 . OOD 
 
 . DX 1 
 
 4 
 
 T"5" 
 
 fi4.9 
 
 TIT 
 
 . 704 
 
 A 
 
 . 729 
 
 
 . ^yu 
 
 6 
 
 CI 
 
 . olO 
 
 a 
 •ft 
 
 ft77 
 
 A 
 
 009 
 
 
 . ybo 
 
 A 
 
 
 
 6.04:9" 
 
 9 
 
 Q (\nAI< 
 y . 074: 
 
 1 1 
 
 .136 
 
 1 0 
 
 .161 
 
 6" 
 
 .222 
 
 1 1 
 
 .247 
 
 
 .309 
 
 9" 
 
 .334 
 
 •A 
 
 .395 
 
 A 
 
 .420 
 
 A 
 A 
 
 .482 
 
 2 
 
 .507 
 
 .568 
 
 A 
 
 .593 
 
 A 
 
 .654 
 
 4 
 
 TU 
 
 .679 
 
 
 .741 
 
 5 
 T2" 
 
 .766 
 
 A 
 
 .827 
 
 A 
 
 .852 
 
 A 
 
 .914 
 
 7 
 
 1 2 
 
 .939 
 
 A 
 
 7.000" 
 
 8 
 TT 
 
 10.025" 
 
 n 
 
 .087 
 
 9 
 T2" 
 
 .112 
 
 H 
 
 .173 
 
 1 0 
 1 2 
 
 .198 
 
 7" 
 
 .259 
 
 
 .284 
 
 A 
 
 .346 
 
 10" 
 
 .371 
 
 A 
 
 .432 
 
 1 
 
 .457 
 
 A 
 A 
 
 .519 
 
 A 
 
 .544 
 
 .605 
 
 3 
 
 .630 
 
 A 
 
 .692 
 
 4 
 
 .717 
 
93 
 
 TABLE OF AUSTRIAN AND ENGLISH mCHES—ConUnued. 
 
 AUSTBIAN. 
 
 English. 
 
 Austrian. 
 
 English. 
 
 AUSTEIAN. 
 
 English. 
 
 1' = 
 
 2' 
 
 3' = 
 
 1' 0.445" 
 2' 0.891" 
 3' 1.336" 
 
 4' = 
 5' = 
 6' = 
 
 4' 1.782" 
 5' 2.227" 
 6' 2.673" 
 
 7' = 
 8' = 
 9' = 
 
 7' 3.118" 
 8' 3.564" 
 9' 4.010" 
 
 Austrian Inches 
 
 English Inches 
 
 Austrian Inches 
 
 English Inches 
 
 IN TWELFTHS. 
 
 IN decimals. 
 
 in twelfths. 
 
 IN decimals. 
 
 10" 
 
 10 . 803" 
 
 11" ^ 
 
 11.667" 
 
 
 .889 
 
 4 
 
 1 2 
 
 .754 
 
 .976 
 
 5 
 
 1 -2 
 
 .840 
 
 
 11.062" 
 
 e 
 
 T2" 
 
 .926 
 
 
 .149 
 
 7 
 T2" 
 
 12.013" 
 
 1 0 
 
 T3" 
 
 .235 
 
 8 
 T2- 
 
 .099 
 
 1 1 
 
 .322 
 
 9 
 T2- 
 
 .186 
 
 11" 
 
 .408 
 
 1 0 
 1 2 
 
 .272 
 
 1 
 
 .494 
 
 1 1 
 
 1 2 
 
 359 
 
 
 .581 
 
 12'^ 
 
 12^44556" 
 
 English Inch 
 
 IN decimals. 
 
 .031 
 .062 
 .093 
 .125 
 .156 
 .187 
 .218 
 .250 
 .281 
 .312 
 .343 
 .375 
 .406 
 .437 
 .468 
 .500 
 
 English Inch 
 IN thirty-seconds. 
 
 1 
 
 y-g- 
 
 3 
 T3" 
 1 
 
 ¥ 
 5 
 
 "Sir 
 
 3 
 "TB" 
 
 i 
 9 
 
 4 
 
 T'B" 
 .15 
 
 1 
 
 English Inch 
 in decimals. 
 
 .531 
 .562 
 .593 
 .625 
 ,656 
 ,687 
 .718 
 ,750 
 .781 
 ,812 
 ,843 
 ,875 
 .906 
 ,937 
 .968 
 
 English Inch 
 in thirty-seconds. 
 
 T"5- 
 
 5 
 "S" 
 2 1 
 
 H 
 
 A 
 4 
 
 1 3 
 T'S" 
 
 2 7 
 
 2 9 
 
 1 a 
 
 TS 
 
 3J, 
 
 3 2 
 
I N D 
 
 E X. 
 
 PAGE 
 
 Adjusting passes 7, 12 
 
 Angle iron 53 
 
 draughting passes for 66 
 
 ways of rolling 67 
 
 arrangement for ^roo«?es for. . . 68 
 
 rolls for 68-70 
 
 Apron, vid. Bearing bar — 
 
 Arrangement of passes 13 
 
 Bearing bar 18 
 
 Body-fillet 4 
 
 Box pass 4, 9, 27 
 
 Chairs, pile for 49 
 
 European form of 73 
 
 Channel iron 71, 72 
 
 Cinder plate , 27 
 
 Closed passes 5 
 
 Collar — defined 3 
 
 Collar rolls (so called) 79 
 
 Cross ends, vid. Rail ends 
 
 Deck beams, vid. T iron 
 
 Diamond passes 10, 36 
 
 Drag rolls 2 
 
 Draw or Draught, defined 14 
 
 amount of. 22 
 
 of Gothic passes 23 
 
 for Styrian iron 26 
 
 of box and flat passes 29 
 
 of diamond passes 38 
 
 of oval passes 40, 42 
 
 of octagon passes 47 
 
 diflference of in various parts 
 
 of passes 51-54 
 
 heavy causes fins 54 
 
 limited by small rolls 56 
 
 Drawing passes 11 
 
 Eccentric passes 7 
 
 Edging passes 6, 12 
 
 PAGE 
 
 End fillets 4 
 
 End roUs, vid. Overhang rolls 
 
 Felly iron, rolls for 73, 74 
 
 Fillets, defined 4 
 
 width of 24, 30 
 
 rounded off 24-25 
 
 Fins, prevention of. 55, 58, 59, 76, 77 
 
 Finishing passes 11 
 
 Fish-belly rails, passes for 7 
 
 Fish plates, rolls for 74 
 
 Flat passes 9, 15, 19, 28, 34 
 
 Flatting passes 7, 12, 15 
 
 Former, defined 3 
 
 must not cut bar 55 
 
 must fit grooves closely 87 
 
 Girders, heavy 60 
 
 Gothic passes 4, 9, 19, 38, 36 
 
 Groove, defined 2 
 
 Guards 18, 54 
 
 keep sharp 55 
 
 hanging 27 
 
 Guides 18, 54 
 
 for step rolls 85 
 
 Half open passes 6 
 
 Hanging guards. . 27 
 
 Heat, influence of on " spread" of iron 16 
 
 Hexagon passes 10 
 
 Hoop iron 16, 30 
 
 Housing (train) defined 1 
 
 (lathe), vid. Steady rest 
 
 I beams, piles for 48, 49 
 
 rolls for 72 
 
 Intermittent passes 9 
 
 Journals, to take thrust 54 
 
 Laying out rolls from given passes. . . 81 
 
i v5 T ( ; ■ ■ 
 
 Neck.'. 1 
 
 method of turning up 83 
 
 Nip.0jf;*f>ae3 , . 34 
 
 of Targe rolls 57 
 
 Octagon passes 10, 46 
 
 Open passes 4, 18 
 
 Oval passes 10, 40 
 
 Overhang rolls 79 
 
 Pass, defined 2 
 
 Pile, formation of for I beams 48 
 
 for chairs with lips 49 
 
 ■ Pitch line 17 
 
 Plate rolls 2 
 
 Play between rolls 5 
 
 Pods 1 
 
 Polishing rolls 2 
 
 Puddle roUs 25, 30, 31, 33 
 
 Rail ends, treatment of for pile 62 
 
 roll for flatting down 63 
 
 Roll, defined 1 
 
 body of. 1 
 
 small compared with largfe .... 56 
 
 method of casting 86 
 
 laying out from given passes. . 31 
 
 Roughing passes 12 
 
 Round passes 10, 44 
 
 Shapes 10, 47 
 
 rolls for small 77 
 
 Shaping passes 11 
 
 Shrinkage 28 
 
 Side pressure 15 
 
 Side thrust, occasioned by unequal 
 
 draw 53 
 
 96 
 
 PAGB 
 
 Side thrust obviated by flatting 
 
 passes 53, 64 
 
 Slitting mill 78 
 
 Speed of large and small rolls 57 
 
 Spike rod 9, 55, 74 
 
 • ■ rolls for 75-77 
 
 Spiral passes 8 
 
 Step rolls 2, 12, 35 
 
 Steady rests 83-85 
 
 Strength of rolls affected by passes. . . 26 
 affected by their own length. . 56 
 
 Templates for passes 82 
 
 T iron, rolls for 70 
 
 T rail, passes 53 
 
 rolls for iron 58-60, 62 
 
 rolls for Bessemer steel 60, 61 
 
 Three high train 27, 38 
 
 vid. also Translator's Preface. 
 
 Tools of chilled iron 85 
 
 with movable ends 86 
 
 Train, defined 13 
 
 Tyre mill 80 
 
 . horizontal 80 
 
 vertical 81 
 
 Tyre rolls, Silesian 64 
 
 Stj'rian 65 
 
 for disc wheels. 65 
 
 Tubes, rolls for 81 
 
 Universal miU 30, 81 
 
 Wash heat 61 
 
 Welding passes 10 
 
 Width of passes 14, 80 
 
 Wire rod 36, 43 
 
 Work on iron in passes 48 
 
Scientific Books^ 
 
 RANCIS' (J. B.) Hydraulic Experiments. Lowell Hydraulic Ex- 
 
 periments — being a Selection from Experiments on Hydraulic 
 Motors, on the Flow of Water over Weirs, and in Open Canals of 
 Uniform Rectangular Section, made at Lowell, Mass, By J. B. 
 Francis, Civil Engineer, Second edition, revised and enlarged, in- 
 cluding maliy New Experiments on Gauging Water in Open Canals, 
 and on the Flow through Submerged Orifices and Diverging Tubes. 
 With 23 copperplates, beautifully engraved, and about 100 new 
 pages of text, i vol., 4to, Cloth. I15. 
 
 M;o8t of the practical rules given in the books on hydraulics have been determined from ex- 
 peri ments made in other countries, with insufficient apparatus, and on such a minute scale, that 
 in applying them to the large operations arising in practice in this country, the engineer cannot 
 but doubt their reliable applicability. The parties controlling the great vs'ater-power furnished 
 by t he Merrimack River at Lowell, Massachusetts, felt this so keenly, that they have deemed it 
 necessary, at great expense, to determine anew some of the most important rules for gauging 
 the flow of large streams of water, and for this purpose have caused to be made, with great care, 
 sevural series of experiments on a large scale, a selection from which are minutely detailed in 
 this volume. 
 
 The work is divided into two parts — Part I., on hydraulic motors, includes ninety-two experi- 
 ments on an improved Fourneyron Turbine Water- Wheel, of about two hundred horse-power, 
 with rules and tables for the construction of similar motors : — Thirteen experiments on a model 
 of a centre- vent water-wheel of the most simple design, and thirty-niue experiments on a centre- 
 vent water-wheel of about two hundred and thirty horse-power. 
 
 Part II. includes seventy-four experiments made for the purpose of determining the form of 
 the formula for computing the flow of water over weirs; nine experiments on the eifect of back- 
 water on the flow over weirs ; eighty-eight experiments made for the purpose of determining 
 the formula for computing the flow over weirs of regular or standard forms, with several tables 
 of comparisons of the new formula with the results obtained by former experimenters ; five ex- 
 periments on the flow over a dam in which the crest was of the same form as that built by the 
 Essex Company across the Merrimack River at Lawrence, Massachusetts ; twenty-one experi- 
 ments on the eff'ect of observing the depths of water on a weir at difi'erent distances from the 
 weir ; an extensive series of experiments made for the purpose of determining rules for gaug- 
 ing streams of water in open canals, with tables for facilitating the same ; and one hundred and 
 one experiments on the discharge of water through submerged orifices and diverging tubes, the 
 whole being fully illustrated by twenty-three double plates engraved on copper. 
 
 In 1855 the proprietors of the Locks and Canals on Merrimack River, at whose expense most 
 of the experiments were made, being willing that the public should share the benefits of the 
 scientific operations promoted by them, consented to the publication of the first edition of this 
 work, which contained a selection of the most important hydraulic experiments made at Lowell 
 up to that time. In this second edition the principal hydraulic experiments made there, subse- 
 quent to 1855, have been added, including the important series above mentioned, for determin- 
 ing rules for the gauging the flow of water in open canals, and the interesting series on the flow 
 through a submerged Venturi's tube, in which a larger flow was obtained than any we flnd re- 
 corded. 
 
 'RANCIS (J. B.) On the Strength of Cast-iron Pillars, with Tables 
 for the use of Engineers, Architects, and Builders. By James B, 
 Francis, Civil Engineer, i vol., 8vo. Cloth. $2. 
 
24 
 
 D. Van Nostrand's Publications. 
 
 HOLLEY'S RAILWAY PRACTICE. American and European 
 Railway Practice, in the Economical Generation of Steam, in- 
 cluding the materials and construction of Coal-burning Boilers, 
 Combustion, the Variable Blast, Vaporization, Circulation, Super- 
 heating, Supplying and Heating Feed-water, &c., and the adaptation 
 of Wood and Coke-burning Engines to Coal-burning ; and in Per- 
 manent Way, including Road-bed, Sleepers, Rails, Joint Fastenings, 
 Street Railways, &c., &c. By Alexander L. Holley, B. P. With 
 77 lithographed plates, i vol., folio. Cloth. $12. 
 
 " This is an elaborate treatise by one of our ablest civil engineers, on the construction and use 
 of locomotives, with a few chapters on the building of Eailroads. * * * All these subjects 
 are treated by the author, who is a first-class railroad engineer, in both an intelligent and intelli- 
 gible manner. The facts and ideas are well arranged, and presented in a clear and simple style, 
 accompanied by beautiful engravings, and we presume the work wiW be regarded as indispens- 
 able by all who are interested in a knowledge of the construction of railroads and rolling stock, 
 or the working of locomotives." — Sdentijic American. 
 
 HENRICI (OLAUS). Skeleton Structures, especially in their Appli- 
 cation to the Building of Steel and Iron Bridges, By Olaus 
 Henrici. With folding plates and diagrams, i vol., 8vo. Cloth. 
 $3. 
 
 WHILDEN (J. K.) On the Strength of Materials used in En- 
 gineering Construction. By J. K. Whilden. i vol., i2mo. 
 Cloth. $2. 
 
 " We find in this work tables of the tensile strength of timber, metals, stones, wire, rope, 
 hempen cable, strength of thin cylinders of cast-iron ; modulus of elasticity, strength of thick 
 cylinders, as cannon, &c., effects of reheating, &c., resistance of timber, metals, and stone to 
 crushing; experiments on brick- work; strength of pillars; collapse of tube; experiments on 
 punching and shearing ; the transverse strength of materials ; beams of uniform strength ; table 
 of coefficients of timber, stone, and iron ; relative strength of weight Ih cast-iron, transversa 
 strength of alloys; experiments on wrought and cast-iron beams: lattice girders, trussed cast- 
 iron girders ; deflection of beams ; torsional strength and torsional elasticity."— ^/?i«rjca» Ar- 
 tisan. 
 
 C AMPIN (F.) On the Construction of Iron Roofs. A Theoretical 
 and Practical Treatise. By Francis Campin. With wood-cuts and 
 plates of Roofs lately executed. Large 8vo. Cloth. $3. 
 
 BROOKLYN WATER-WORKS AND SEWERS. Containing a 
 Descriptive Account of the Construction of the Works, and also 
 Reports on the Brooklyn, Hartford, Belleville, and "Cambridge 
 Pumping Engines. Prepared and printed by order of j.e Board of 
 Water Commissioners. With illustrations. i vol., folio. Cloth. 
 $15. 
 
 ROEBLING (J. A.) Long and Short Span Railway Bridges. By 
 John a. Roebling, C. E. Illustrated with large copperplate en- 
 gravings of plans and views. In press. 
 
 CLARKE (T. C.) Description of the Iron Railway Bridge across 
 the Mississippi River at Quincy, Illinois. By Thomas Curtis 
 Clarke, Chief Engineer. Illustrated with numerous lithographed 
 plans. I vol., 8vo. Cloth. $7.50. 
 
Scientific Boohs. 
 
 25 
 
 ILLIAMSON (R. S.) On the Use of the Barometer on Surveys 
 and Reconnaissances. Part I. Meteorology in its Connection 
 with Hypsometry. Part II. Barometric Hypsometry. By R. S. 
 Williamson, Bvt. Lieut. -Col. U. S. A., Major Corps of Engineers. 
 With Illustrative Tables and Engravings. Paper No. 15, Professional 
 Papers, Corps of Engineers, i vol., 4to. Cloth. $15. 
 
 " San Francisco, Cal., Feb. 27, 1867. 
 " Gen. A. A. Httmphbets, Chief of Engineers, U. S. Army: 
 
 " GbneraI/— I have the honor to submit to you, in the following pages, the results of my in- 
 vestigations in meteorology and hypsometry, made with the view of ascertaining how far the 
 barometer can be used as a reliable instrument for determining altitudes on extended lines of 
 Buirvey and reconnaissances. These investigations have occupied the leisure permitted me from 
 miy professional duties during the last ten years, and I hope the results will be deemed of suffl- 
 ci ent value to have a place assigned them among the printed professional papers of the United 
 Sttates Corps of Engineers. Very respectfully, your obedient servant, 
 
 "E. 8. WILLIAMSON, 
 "Bvt. Lt.-Col. U. S. A., Major Corps of U. S. Engineers.*' 
 
 TUNNER (P.) A Treatise on Roll-Turning for the Manufacture of 
 Iron. By Peter Tunner. Translated and adapted. By John B. 
 Pearse, of the Pennsylvania Steel Works. With numerous engrav- 
 ings and wood-cuts. In press. 
 
 SHAFFNER (T. P.) Telegraph Manual. A Complete History and 
 Description of the Semaphoric, Electric, and Magnetic Telegraphs 
 of Europe, Asia, and Africa, with 625 illustrations. By Tal. P. 
 Shaffner, of Kentucky. New edition, i vol., 8vo. Cloth. 850 pp. 
 $6.50. 
 
 MINIFIE (WM. ) Mechanical Drawing. A Text-Book of Geomet- 
 rical Drawing for the use of Mechanics and Schools, in which 
 the Definitions and Rules of Geometry are familiarly explained ; the 
 Practical Problems are arranged, from the most simple to the more 
 complex, and in their description technicalities are avoided as much 
 as possible. With illustrations for Drawing Plans, Sections, and 
 Elevations of Buildings and Machinery ; an Introduction to Isomet- 
 rical Drawing, and an Essay on Linear Perspective and Shadows. 
 Illustrated with over 200 diagrams engraved on steel. By Wm 
 MiNiFiE, Architect. Seventh edition. With an Appendix on the 
 Theory and Application of Colors, i vol., 8vo. Cloth. $4. 
 
 ' It is the best work on Drawing that we have ever seen, and is especially a text-book of Geo- 
 metrical Drawing for the use of Mechanics and Schools. No young Mechanic, such as a Ma- 
 chinist, Engineer, Cabinet-Maker, Millwright, or Carpenter should be without \t.''—ScientiJie 
 American. 
 
 " One of the most comprehensive works of the kind ever published, and cannot but possess 
 great value to builders. The style is at once elegant and substantial." — Pennsylvania Inquirer. 
 
 " Whatever is said is rendered perfectly intelligible by remarkably well-executed diagrams on 
 Bteel, leaving nothing for mere vague supposition ; and the addition of an introduction to iso- 
 metrical drawing, linear perspective, and the projection of shadows, winding up with a useful 
 index to technical terms." — Glasgow Mechanics'' Journal. 
 
 The British jvernment has authorized the use of this book in their schools of art at 
 Somerset House, London, and throughout the kingdom. 
 
 MINIFIE (WM.) Geometrical Drawing. Abridged from the octavo 
 edition, for the use of Schools. Illustrated with 48 steel plates. 
 Fifth edition, i vol., i2mo. Half roan. $1.50. 
 
 "It is well adapted as a text-book of drawing to be used in our High Schools and Academies 
 Where this useful branch of the fine arts has been hitherto too much neglQcted."— Boston JoumOr 
 
26 
 
 D. Van JSfostrand' s Publications. 
 
 PEIRCE'S SYSTEM OF ANALYTIC MECHANICS. Physical 
 and Celestial Mechanics, by Benjamin Peirce, Perkins Professor 
 of Astronomy and Mathematics in Harvard University, and Con- 
 sulting Astronomer of the American Ephemeris and Nautical Al- 
 manac. Developed in four systems of Analytic Mechanics, Celestial 
 Mechanics, Potential Physics, and Analytic Morphology, i vol., 
 4to. Cloth. $io. 
 
 GILLMORE. Practical Treatise on Limes, Hydraulic Cements, and 
 Mortars. Papers on Practical Engineering, U. S. Engineer De- 
 partment, No. 9, containing Reports of numerous experiments con- 
 ducted in New York City, during the years 1858 to 1861, inclusive. 
 By Q. A. GiLLMORE, Brig. -General U. S. Volunteers, and Major U. 
 S. Corps of Engineers. With numerous illustrations. One volume, 
 octavo. Cloth. $4. 
 
 ROGERS (H. D.) Geology of Pennsylvania. A complete Scien- 
 tific Treatise on the Coal Formations. By Henry D. Rogers, 
 Geologist. 3 vols., 4to., plates and maps. Boards. $30.00. 
 
 BURGH (N. P.) Modern Marine Engineering, applied to Paddle 
 and Screw Propulsion. Consisting of 36 colored plates, 259 
 Practical Woodcut Illustrations, and 403 pages of Descriptive Matter, 
 the whole being an exposition of the present practice of the follow- 
 ing firms : Messrs. J. Penn & Sons ; Messrs. Maudslay, Sons, & 
 Field ; Messrs. James Watt & Co. ; Messrs. J. & G. Rennie ; Messrs. 
 R. Napier & Sons ; Messrs. J. & W. Dudgeon ; Messrs. Ravenhill 
 & Hodgson ; Messrs. Humphreys & Tenant ; Mr. J. T. Spencer, 
 and Messrs. Forrester & Co. By N. P. Burgh, Engineer. In one 
 thick vol., 4to. Cloth. $30.00. Half morocco. $35.00. 
 
 KING. Lessons and Practical Notes on Steam, the Steam-Engine, 
 Propellers, &c. , &c., for Young Marine Engineers, Students, 
 and others. By the late W. R. King, U. S. N. Revised by Chief- 
 Engineer J. W. King, U. S. Navy. Ninth edition, enlarged. 8vo. 
 Cloth. $2. 
 
 WARD. Steam for the Million. A Popular Treatise ©n Steam and 
 its Application to the Useful Arts, especially to Navigation. By 
 J. H. Ward, Commander U. S. Navy. New and revised edition. 
 I vol., §vo. Cloth. $1. 
 
 WALKER. Screw Propulsion. Notes on Screw Propulsion, its 
 Rise and History. By Capt. W. H, Walker, U. S. Navy, i 
 vol., 8vo. Cloth. 75 cents. 
 
 THE STEAM-ENGINE INDICATOR, and the Improved Mano- 
 meter Steam and Vacuum Gauges : Their Utility and Application. 
 By Paul Stillman. New edition, i vol., i2mo. Flexible cloth. 
 $1. 
 
 I SHERWOOD. Engineering Precedents for Steam Machinery. Ar- 
 ranged in the most practical and useful manner for Engineers. Bj 
 B. F. IsHERwooD, Civil Engineer U. S. Navy. With illustration* 
 Two volumes in one. 8vo. Cloth. $2.50. 
 
Scientific Boohs. 
 
 27 
 
 OOK'S METHOD OF COMPARING THE LINES AND 
 
 -L DRAUGHTING VESSELS PROPELLED BY SAIL OR 
 STEAM, including a Chapter on Laying off on the Mould-Loft 
 Floor. By Samuel M. Pook, Naval Constructor. i vol., 8vo. 
 With illustrations. Cloth. $5. 
 
 SWEET (S. H. ) Special Report on Coal ; showing its Distribution, 
 Classification and Cost delivered over different routes to various 
 points in the State of New York, and the principal cities on the 
 Atlantic Coast. By S. H. Sweet. With maps, i vol., 8vo. Cloth. 
 
 ALEXANDER (J. H.) Universal Dictionary of Weights and Meas- 
 ures, Ancient and Modern, reduced to the standards of the United 
 States of America. By J. H. Alexander. New edition, i vol., 
 8vo. Cloth. $3.50. 
 
 " As a standard work of reference this book should be in every library ; it is one which wo 
 have long wanted, and it will save us much trouble and vQSQ&cch.y—ScientiJic American. 
 
 CRAIG (B. F. ) Weights and Measures. An Account of the Deci- 
 mal System, with Tables of Conversion for Commercial and Scien- 
 tific Uses. By B. F. Craig, M. D. i vol., square 32mo. Limp 
 cloth. 50 cents. 
 
 " The most lucid, accurate, and useful of all the hand-books on this subject that we have yet 
 seen. It gives forty-seven tables of comparison between the English and French denominations 
 of length, area, capacity, weight, and the centigrade and Fahrenheit thermometers, with clear 
 instructions how to use them ; and to this practical portion, which helps to make the transition 
 CSS easy as possible, is prefixed a scientific explanation of the errors in the metric system, and 
 how they may be corrected in the laboratory."— iVaiiiow. 
 
 BAUERMAN. Treatise on the Metallurgy of Iron, containing 
 outlines of the History of Iron manufacture, methods of Assay, 
 and analysis of Iron Ores, processes of manufacture of Iron and 
 Steel, etc., etc. By H. Bauerman. First American edition. Re- 
 vised and enlarged, with an appendix on the Martin Process for 
 making Steel, from the report of Abram S. Hewitt. Illustrated 
 with numerous wood engravings. i2mo. Cloth. $2. 5b. 
 
 " This is an important addition to the stock of technical works published in this country. It 
 embodies the latest facts, discoveries, and processes connected with the manufacture of iron 
 and steel, and should be in the hands of every person interested in the subject, as well as in all 
 technical and scientific libraries."— ^ci«n<i/fo American. 
 
 HARRISON. Mechanic's Tool Book, with practical rules and sug- 
 gestions, for the use of Machinists, Iron Workers, and others.^ 
 By W. B. Harrison, associate editor of the "American Artisan." 
 Illustrated with 44 engravings. i2mo. Cloth. $2.50. 
 
 " This work is specially adapted to meet the wants of Machinists and workers in iron gener- 
 ally It is made up of the work-day experience of an inteUigent and ingenious mechanic, who 
 had the faculty of adapting tools to various purposes. The practicability of his plans and sug- 
 gestions are made apparent even to the unpractised eye by a series of well-executed wood en- 
 gravings."— PAiZacJeZpAia Inquirer. 
 
 $3- 
 
28 
 
 D. Van Nostrand's Publications. 
 
 PLYMPTON. The Blow-Pipe : A System of Instruction in its prac- 
 tical use, being a graduated course of Analysis for the use of 
 students, and all those engaged in the Examination of Metallic 
 Combinations. Second edition, with an appendix and a copious 
 index. By George W. Plympton, of the Polytechnic Institute, 
 Brooklyn. i2mo. Cloth. $2. 
 
 This manual probably has no superior in the English language as a text-book for beginners, 
 or as a guide to the student working without a teacher. To the latter many illustrations of the 
 utensils and apparatus required in using the blow-pipe, as well as the fully illustrated descrip- 
 tion of the blow-pipe flame, will be especially serviceable."— iVe?i; York Teacher. 
 
 NUGENT. Treatise on Optics : or. Light and Sight, theoretically 
 and practically treated ; with the application to Fine Art and In- 
 dustrial Pursuits. By E. Nugent. With one hundred and three 
 illustrations. i2mo. Cloth. $2. 
 
 " This book is of a practical rather than a theoretical Mnd, and is designed to afford accurate 
 and complete information to all interested in applications of the science."— ^0MW(i TcUile. 
 
 O ILVERSMITH (Julius). A Practical Hand-Book for Miners, Met- 
 ^ allurgists, and Assayers, comprising the most recent improvements 
 
 in the disintegration, amalgamation, smelting, and parting of the 
 
 Precious Ores, with a Comprehensive Digest of the Mining Laws. 
 
 Greatly augmented, revised, and corrected. By Julius Silversmiih. 
 
 Fourth edition. Profusely illustrated, i vol., i2mo. Cloth. $3. 
 
 LOUGH. The Contractors' Manual and Builders' Price-Book. By 
 ^ A. B. Clough, Architect, i vol., i8mo. Cloth. 75 cents. 
 
 BRUNNOW. Spherical Astronomy. By F. Brunnow, Ph. Dr. 
 Translated by the Author from the Second German edition, i 
 vol., 8vo. Cloth, I6.50. 
 
 r'^HAUVENET (Prof. Wm.) New method of Correcting Lunar Dis- 
 ^ tances, and Improved Method of Finding the Error and Rate of a 
 
 Chronometer, by equal altitudes. By Wm. Chauvenet, LL. D. i 
 
 vol., 8vo. Cloth. $2. 
 
 A SYNOPSIS OF BRITISH GAS LIGHTING, comprising the 
 -^V essence of the "London Journal of Gas Lighting" from 184Q to 
 1868. Arranged and executed by James R. Smedberg, C. E. of the 
 San Francisco Gas Works. Issued only to subscribers. 4to. Cloth. 
 $15.00 In press. 
 
 AS WORKS OF LONDON. By Zerah Colburn. i2mo. Boards. 
 60 cents. 
 
 G 
 H 
 
 EWSON. Principles and Practice of Embanking Lands frcm 
 River Floods, as applied to the Levees of the Missi.ssippi. 3y 
 William Hewson, Civil Engineer, i vol., 8vo. Cloth. $2. 
 
 " This is a valuable treatise on the principles and practice of embanking lands from river 
 floods, as applied to Levees of the Mississippi, by a highly intelligent and experienced engineer. 
 The author says it is a first attempt to reduce to order and to rule the design, execution, tnd 
 'aeasurement of the Levees of the Mississippi. It is a most ueeiid and needk>v. contribution to 
 '.ientiflc literature."- PAi^ade/pAia Evening Journal.