[-UBRARY0/ ; OF-CAUFO^ I 1 ^ % RARY0A flMIW^ S 1 t 1 i %. ari . UNIVER% 3 i I t OKAllF(% -^ T? RAILROAD ENGINEER'S PRACTICE, A SHORT BUT COMPLETE DESCRIPTION OF THE DUTIES OF THE YOUNG ENGINEER IN PRE- LIMINARY AND LOCATION SURVEYS, AND IN CONSTRUCTION. BY THOMAS M. CLEEMANN, A. M., C. NEW YORK : GEORGE H. FROST, PUBLISHER. 1880. COPYRIGHT : THOMAS M. CLEEMANN. 1880. ITKIN & PROUT, PRINTERS, No. 13 BARCLAY STREET, New York. ERRATA. On page 14, line 4, for 1^ read %. On page 26, line 2 from bottom, for J read 1^. On page 52, lines 7, 9 and 11, for W read VF'f in lineg and 10 the lirintinrr ia nnt nlrn-m i-n nnU ... - *i c < ERRATA. On p. 39, 1. 8 from bottom, for "less" read "more," and the next sentence may then be left out. On p. 49, 1. 2 from top, ' ' The maximum compression on 1X2 ?i 3 D B = W sec. 6 " should have added to it, " W 2X n 2 sec. 6 if a plus quantity ; if it is a minus quantity, there will be no compression on D Z?." 2X3 On p. 49, 1. 7 from top, " The maximum tension on B C 2n n 5 W sec. " should have added to it " Wsec. 6 if a plus 2 quantity." On p. 49, 1. 8 from top, " The maximum compression on 3X4 n1 C I = W sec. " should have added to it " W 2n 2 sec. Q if a plus quantity." 4X5 On p. 49, 1. 12 from top, " The maximum tension on / L = 2n n Q W sec. 6 " should have added to it " Wsec. if a plus 2 quantity." ATKW & PBOTTT, PRINTERS, No. 12 BARCLAY STREET, New York. ERRATA. On page 14, line 4, for 1^ read %. /" On page 26, line 2 from bottom, for J. read 1^. On page 52, lines 7, 9 and 11, for W read W; in lines 8 and 10 the printing is not clear; in each one the first term of the second member of the equation is W, and the sec- ond term W. On page 54, lines 9 and 11, place a figure 2 before W and W. On page 55, line 8 from bottom, after " total width " read " or depth (whichever is the least)." On page 57, in the lower figure, the letter L is left out; it should be just above the letter A. the Chief Engineer under whom the writer began the practice of his profes sion, on the Pennsylvania Railroad, and whose uniform kindness and interest in his welfare have been continual causes of pleasure and gratitude, this book is respectfully dedicated by T. M. C. ATKIN & PKOTJT, PRINTERS, No. 12 BARCLAY STREET, New York. TO W. H. WILSON, Esq., the Chief Engineer under whom the writer began the practice of his profes- sion, on the Pennsylvania Railroad, and whose uniform kindness and interest in his welfare have been continual causes of pleasure and gratitude, this book is respectfully dedicated by T. M. C. TF 5.oo C54 PREFACE. The present work is intended to fill a want that the writer him- self acutely felt on beginning his professional career. There were many points of practice of which he could only get information by observation and experience, involving a loss of time, which s might have been saved to him, could he have referred to some / book which would have told him how certain problems had been solved by other engineers, and which would have prepared him , the better to observe the methods pursued by those with whom ^ he was thrown in executing their work. Some of those who approve the intention of the book, may ^? think that too much detail has been entered into, and that much ' ;j that is specified can be more quickly and better learned in the VvJ field. To such it is replied that it is extremely difficult to tell ^ the proper point at which to stop in minute description, and that if ^ the author has erred in minuteness, it is at least a fault on the . proper side ; he will be satisfied if he has left nothing essential ^ out. In regard to the methods of organizing parties, keeping field X books, staking out, etc., there are very great differences among &T engineers; the author has merely given what he individually con- sidered the best. ^\ The subject of iron bridges has not been touched upon, be- v^cause it would add very considerably to the bulk of the book. If, however, this first venture of the author should prove a success, he hopes to devote a special book to the subject of bridges, in hich an attempt will be made to treat of the various details that are employed, in an exhaustive manner, as well as to discuss the strains in the different "skeleton " structures. The author has freely quoted from the "Engineer's Pocket Book," of Mr. Trautwine, which he hopes will be in the library of every engineer who may possess the present work. He con- siders no other book to have appeared which supplies so well a constant want of the engineer at all stages of his career, and he is happy to be able to state that Mr. Trautwine himself gave him permission to extract from it what he desired for the present book. He has tried to give credit to those from whom he has ob- tained information, but as much of his material lias been obtained from conversations with other engineers, notes of which were made for his own convenience only, at a time when he did not expect to put them in book form, he cannot always recall to whom he is indebted. If any of his friends, however, remember giving him any practical rules which are not acknowledged, he will be obliged if they will notify him of the fact. 340 SOUTH TWENTY-FIRST STREET, ) PHILADELPHIA, October, 1879. ) TABLE OF CONTENTS. PRELIMINARY SURVEY. PAGE. Inspection by Chief Engineer 7 Method Pursued by Principal Assistant Engineers 7 Organization of Parties 7 Starting a Survey 7 Duties of Transitman, Leveller and Topographer 8 Convenient Form of Clinometer 10 Form of Transit Book 10 Topographer's Duties 11 The Paper Location 13 Form for Excavation and Embankment 13 Slopes of Cuts and Fills 14 LOCATION. Organization of Party 14 Starting the Location 14 Problems m Curves 14 Transit Points 23 Form of Field-Book 23 Div ision of Line into Sections 23 Letting the Contracts 23 CONSTRUCTION. Principal Assistant Engineer ; his Party and Duties 24 Retracing Line 24 Guarding " Plugs " 24 Form of Note-Book 25 Setting Slope-Stakes 26 Form of Field-Book 27 Calculating Cross-Section Areas 28 Estimates of Work 29 CULVERTS. Finding Water-Way 29 Box CULVERTS. How to Lay Out 30 Proper Size 31 4 PAGE. OPEN CULVERTS 31 CATTLE-GUARDS 31 OPEN PASSAGE-WAYS 31 STONE ARCHES. Formulae 34 Centring 36 RETAINING WALLS. Formulae 37 TUNNELS. The," Heading" , 40 A Summit 41 Shafts 41 Arching 41 Form of Intrados 41 Timbering 43 Size of Excavation 43 Blasts in Excavation 43 BRIDGES. Formulae 44 to 54 Greatest Variable Load 54 English and American Practice 54 Proper Place for Pin 55 Howe Truss "Keys " 55 Howe Truss Splice 56 Sizes of Timber 57 Camber 58 Brar-ing 58 Superstructure 58 Trestle- Work 59 MASONRY. Contractors' Tricks 60 Rankine's Rule 60 Preparation of Mortar 61 Cement Mixing and Using 61 FOUNDATIONS. Crushing Strains of Stone , 61 Loadj per Square Foot 61 Experiments of Sir Charles Fox and Mr. Leonard. 63 Rip-rapping 62 On Gravel in Water 62 PILE-DRIVING. Formulae 63 Safe Load on Piles... ..64 5 PAGE. Proper Diameter 64 Bearing Power of Discs 65 Through Boulders and Gravel 65 Through Sand 65 Surface Friction of Cast-Iron Cylinders 65 TBACK-LAYING. Re-running Centre Line 65 Rule for Track-1 aying 66 Method of Work : 66 Bending Rails 66 SWITCHES. Formulae 67 Frogs, Turnouts, etc 63 CROSS-TIES How Made and Piled 70 Public Road-crossings 70 RAILS. WATER STATIONS. Use of Water in Engines 71 Gravity Supply 72 Stand-Pipe 72 Tanks 72 Reservoirs 73 Steam and Wind Power 73 COALING STATIONS 73 PASSENGER STATIONS. Description of Cresson Station 73 TELEGRAPH LINE 74 PRELIMINARY SURVEY. The Chief Engineer, from an inspection of the various maps of the country he can obtain, and a personal exami- nation of the ground, decides where it will be necessary to run lines to determine which is the cheapest that can be built, and gives the necessary orders for such lines to the Principal Assistant Engineers. The method pursued by the Principal Assistant Engineer differs according to the character of the country, and the time that can be devoted to the preliminary survey. A quick, rough method of gaining the requisite information for the location will first be given, and afterward, one more exact that pursued on the Bennett's Branch Extension of the Allegheny Valley Railroad. The latter is especially recommended where the means of the company will admit of the more accurate work, and where it may not be a, matter of policy to begin the construction of the road at the earliest possible moment. Each Principal Assistant organizes a party which con- sists as follows : Principal Assistant, Transitman, Leveller, Topographer, Level Rodman, Slope Rodman, Flagman, two Chainmen, three or more Axemen. An Axeman provides a number of stakes in advance and numbers them consecu- tively from 0, shaving off a smooth place for that purpose, and drives the first one usually driven flush with the ground, and called a " plug " at the place indicated by the Prin 1- pal Assistant. The latter then starts ahead with the Flag- man, and the Transitman sets his transit over the first stake or plug. The Principal Assistant, having decided where he wishes to run the line, sets up the flag. The Chainmen instantly begin chaining toward it, the hind one " lining " the head one, and an Axeman driving the stakes, one every 100 feet, in the order of marking. The Transit- man takes his sight, reads only the needle to quarter de- grees, records the reading, and starts off for the flag. On arriving there, he sets up ready for another sight, which the Principal Assistant is ready to give him, by Avaving his handkerchief if the Flagman has not had time to come up. In an open country, the speed of the Chainmen should govern the speed of the party. When there is much underbrush, the Principal Assistant may require several Axemen to clear the way. The Leveller follows the Transitman as closely as possi- ble, taking levels on every stake, and, if necessary, on abrupt intermediate changes of the slope. His Axeman makes " pegs " (or turning points) and cuts down brush obstructing his view. The Topographer follows a day behind the Transitman and Leveller. He is provided with a thin box, with a hinged cover on the end, which serves both as a portfolio and a drawing-board. There should be some oiled cloth fastened to it for keeping the paper dry. The paper is tacked to the board with thumb tacks ; a convenient size of sheet is 21 X 16 inches. The Topographer has obtained at night from the Transitman and Leveller their notes of the day, and plotted the line on a scale of 400 feet to the inch, noting the elevations at the stations. He takes this into the field with him. His Slope Rodman and Axeman go ahead and measure the transverse slopes, laying a rod upon the ground at each station, and upon it a clinometer ; with a tape they measure the distance to where the slope changes, and then measure the new slope and its length, and so on. This is done on each side of the line, and is noted in the Rodman's book in one of two ways : the direc- tion of the slope being indicated either by the signs + and , or by the inclination of the line dividing the numerator from the denominator of a fraction in which the numerator is the angle, and the denominator is the distance. The notes are given to the Topographer, and with the help of the elevations already obtained from the Leveller, he sketches in the contours. To facilitate this, he uses a table which gives the horizontal distance between two contours taken ten feet apart for each degree, as follows : 1 is 573 per 10 feet rig 2 286 " " 3 191 " 4 143 5 114 M 6 95 " 7 81 (t 8 90 71 11 10 57 M 11 51 12 47 M 13 43 M It is well for him to commit this table to memory. He estimates distances beyond those measured by the Rodman, and so puts in distant hills, &c. For this it is convenient to have a pocket sextant. The Rodman runs out to different distances, according to the nature of the ground. If the country is level, he may run out 500 feet on each side of the centre-line, only doing so, perhaps, at intervals of 500 feet, or it may be necessary to run out that distance at every station. If the country is hilly, it maybe sufficient to run out only 100 feet at each station. 1C A convenient form of clinometer is formed of a square board, with a string and bullet : The more exact method is thus described in a private letter written by Mr. A. B. Nichols, in 1870, when he was Principal Assistant on the Bennett's Branch Extension of the Allegheny Valley Railroad : " I always run experimental with the vernier as follows : Go- ing ahead by myself, I select about the spot where I want to 'plug,' and let the Transitman take a sight on me, setting his vernier to the nearest quarter degree (except in special cases). I have the head Chainman carry a sight-staff, and set all the stakes with the transit. The head Chainman then sets the fore-sight plug when he arrives at the end of the sight. I use the needle merely as a check on the vernier. I think it better to set the FORM OF TRANSIT NOTE BOOK. Sta. Angle. Deduced Course. Needle. Remarks. 8 7 turnpike. 6 B 4 L. 6 N. 32% W. N. 32% W. 3 + 30 2 1 R. 1335' N. 26% W. N. 26% W. o N 40 W stakes with the transit, as they are more reliable as references o& location, and in an open country they can be set as fast as 'Abo Leveller can run (beyond which speed there is no use in running), while in a wooded section there is plenty of tune to set th~m 11 while the Axemen are clearing. In thick woods, the Principal Assistant's voice has often to be taken as the guide ahead. The bench marks should be marked with the number of the station immediately preceding, and the distinctive letter of the line. Thus, if there happen to be a bench at 7 + 60 of ' H ' line, it should be marked B. M. | 7 'H.' "In regard to the Topographer's duties, I do not like the system of putting in the topography in the field. It has always been the custom, I believe, to run experimental one day and locate over the next. Mr. J. A. Wilson's method differs some- what from this, and, I think, with reason. Topography put in in the haute that is inevitable in the field, is liable to many errors, and locations made on the previous day's experimental may not suit the country ahead. Mr. Wilson's method is to run all the necessary lines, take all the necessary notes, and then go into office quarters and work the maps up, and make a paper location which may then be run in and modified in the field. "In ' Morrison's Cove ' Mr. Linton took charge of the topo- graphical department, taking the topography notes himself. His instruments were: a pocket compass, mounted on a light tripod, a Locke's level, and a small slope-board. His method of proceed- ing was as follows : "Say that A is a tree on a hill, and B another point on another hilL He would set his compass up at 491, for instance, take the 12 courses to A and B, and measure the vertical angle with his slope- board. He would then proceed to A and take slopes in all direc- tions, and in like manner from B, using his slope-board and level for heights and slopes. Then going to another station, as 500, he would fix the points A and B by other courses and slopes. Hollows can be shown by running a course up and taking slopes to right and left. By that means he could show tbe topog- raphy sometimes a half a mile from the line. I have known his elevations, say at A, to come within a foot of each other at the distance of half mile from the line, deduced from vertical angles taken with the slope-board, as from 491 and 500 ; seldom over two feet difference. "The slope-board is a modification of the square-board and bullet. It will read to quarters of a degree, is furnished with sights,* and is used as follows : " The Assistant Topographer takes his stand at the station, and gives the right angle to the line by means of a right-angle box, or otherwise. The Slope Rodman measures out the horizontal distance with a ten-feet-long pole to change of slope, and sights back on the man at the station, taking a point on the other's person (previously determined) at the same height above ground as his own eyes. He reads the slope, calls it and the distance out, and in the meanwhile the man at the station, be it the Assistant or the other Rodman, checks the slope by sigh'ting on his person. Rodman No. 1 then measures ahead to the next change, while Rodman No. 2 comes up to change No. 1 ; they measure the slope, and so on. The Assistant keeps the books, and should be fur- nished with a ' Jacob's staff ' and compass for taking buildings, and while so engaged the Rodmen can measure the sizes of said building with a tape, or can go on taking slopes, which they afterward report to the Assistant. Slopes should never be esti- mated, except one at the end of a series, and then it should be so marked, and the contours derived from it should be dotted on the map to avoid errors in location. In taking short slopes, one Rod- man can take the right and the other the left of the line, thus facilitating matters." From an inspection of the maps, it will be seen on which routes it is necessary to have paper locations made. A * Mr. Linton's improved slope instrument may now be obtained at mathe- matical instrument makers. 18 paper location is such a line drawn upon the plan as may appear, taken in connection with the profile, to require the least excavation and embankment. The following is an excellent method of obtaining the cheapest location on the preliminary map : Having located a trial line by inspection, a profile is made, and grades assumed and drawn. A hori- zontal plane is supposed to pass through the point on the grade line at each station, and a point, in its line of inter- section with the ground surface opposite the station, is marked with a red point. Having plotted these red points for a sufficient distance, they are connected by a line, which will resemble a contour line, and actually becomes one when the grade is level. The nearer the paper location can be drawn to this line, the less will be the excavation and embankment. If it coincides with it, the line will be a surface line. Having made the locations on such lines as are considered desirable, a new profile is made from an inspection of where the located line cuts the contours, and cross-sections are plotted on a scale of ten feet to the inch, or on Traut- wine's cross-section paper. From these cross-sections, the amounts of excavation and embankment are calculated, and the results embodied in a table of the following form : Sta. Dis- tance. Tfleva. Grade. g *] s AREAS SOLIDS. Excavation. Embank- ment. Rock. Earth. Emb. Rock. Earth. The slopes of the cuts are sometimes assumed as fol- lows : When the slope of the ground is 20 or less, make the slope 1% tot 14 When the slope of the ground is 20 to 35*, make the slope 1 tol. When the slope of the ground is over 35, make a vertical wall. Rock stands at^^ to 1. Embankment is generally taken as sloping 1% to 1. From the calculated amount of excavation and excess of embankment, an estimate is made of the costs of different routes, and it is thus found which lines it will be necessary to actually locate in the field, in order to obtain a closer estimate, or for the purpose of constructing^ LOCATION. The locating party is organized somewhat differently from the preliminary one. We have : Principal Assistant, Transitman, Leveller, Level-Rodman, Front Flagman, Back Flagman, two Chainmen, two or more Axemen. The axeman who drives the stakes now carries tacks, or, better, lath nails, with him, and drives one in the plug at the point where the transitman is to set up. The transit- man uses the vernier entirely, not using the needle, unless as a check on the tangents. The curves are all run on the ground, and the stakes which come upon them, set with the transit. The leveller keeps close up to the transitman, and constantly reports the heights to the Principal Assistant. The tangents are generally fixed by the paper locations, and the usual object is to run a given curve from one end of a tangent, and strike the hill with the point of tangent (P. J!) at a given elevation, viz., thatn the paper location. Other problems will also often arise. The following are the most useful : Problem 1. To change a curve so that it shall come out in a parallel tangent at a given distance from the old tangent, by 15 changing the radius. (From "Haslett & Hackley's Pocket- Book/') A I/ To change the curve A B so that it shall come out at C, * I GO*. I; or otherwise, Degree of curve A C = degree of curve A B T in which n is the number of 100-feet chords in A B. 8 DC Problem 2. To change the origin of a curve so that it may pass through a given point. To move A, the point of compound curvature, so that The curve A B will pass through the point C. Take the distance B C, divide it by A B, and multiply by 57. 3, and we get the difference in deflection CAB, which, divided by the number of stations in A B, gives the difference in deflection ft C 1 per station (or look in the table of natural sines for , from A B which is obtained the angle CAB, which is to be divided by the number of stations). Then take the difference between the de- grees of the curves A B and D A ; then, the difference between the degrees of the curves is to the difference in deflection per % , 16 station as 100 is to the number of feet forward or backward we must go, on the curve A D, to strike the point C. Problem 3. Having located a compound curve terminating in a tangent, it is required to change the point of compound curva- ture so that the curve will terminate in a tangent parallel to the located tangent, at any required distance perpendicular thereto. Divide the required distance between parallel tangents by the difference of radii of the two last branches of the curve. From the cosine of total amount of curvature in the last branch, sub- tract or add this quotient. The remainder, or sum, will be the natural cosine of the amount of curvature required for the last radius. Problem 4. Having located a compound curve terminating in a given tangent, it is required to change the point of compound curvature, and also the length of the last radius, so as to pass through the same terminating point with a given difference in the direction of the tangent. Having the curve H A and the curve A B, with tangent B D, it isrequired to continue the first curve from A to such a point, P, that the tangent at B will have the direction B E. Continue the curve H A to the point P given by the following equation : Cotan. %AMP =^ (cotan. % A NB + cotan. % ) cotan. % a. The curve from P to B is, of course, found by measuring the total deflection angle, and dividing by the number of stations. 17 We can, by means of this problem, connect two curves running toward each other with a third one. Let A and B be points on the respective curves. We wish to continue the curve H A past A to some point, P, from which to run some third curve connect- ing with the other curve at B (the tangent B E being common to the last two curves). Measure the angle G A B = % A N B ; also the angle K B A = Y z A N B + a ; and the distance AB = 2R' sin. % A NB. Then calculate A M Pfrom the above equation ; dividing by the degree of the curve H A, gives the distance A P in stations. From Y z P O B = A NB + a A MP and the distance P B, we can find the degree of the curve P B. Problem 5. To change the radius of a curve so that it will come out in a given tangent. To change the radius of the curve E D so that it will come out in the tangent C H. Having run the curve until the tangent D K is nearly parallel to C H, measure the offsets D G and I K, and the distance G I. iDG\ Calculate G F and then a ( = tan. ^r~p ). We could also measure this angle directly by measuring off M H = D G and taking a sight on M. Then A C = A B + B C = R ver. sin. a. + DG. We then find tin new radius by Problem 1. AC (I = the former total angle minus a). Problem 6. Having located a curve connecting two tangents, 18 it is required to move the middle of the curve any given distance either toward or from the vertex. A Bar AC 1- Problem 7. To change the origin of a curve so that it shall terminate in a tangent parallel to a given tangent at a given dis- tance from it. Let TT\)Q the curve, FA the given tangent, and V T""the parallel tangent. 4 ft" Then T T" = 1- T Problem 8. To find how far back it is necessary to go from the point B, to strike the point C with a curve of given radius ; B C being known. 19 r-^ J 13 ra l*T. x =|/& (2 R b). Problem 9. To draw a tangent to a curve from a point outside of it. A_____; ^_^ vO Sin. BAO ** J sin._^ ~ 2 sin. a sin. D A O sin. B ^1 O Problem 10. To draw a tangent to two curves already located, 1st. When the curves are in opposite directions. C 20 Stop both curves before getting to the tangent points. Ob- serve B AC and A B D, and measure A B. In the triangle ABC, calculate B, C B A and A C B; (A C, A B and CAB being known). In the triangle BCD calculate C D, B D C and D C B; (BC,BD and CB D = CBA + ABD, being known). BDE=BDCEDC 2d. When the curves are in the same direction. This is an ex- tension of Problem 4. Problem 11. To substitute a compound curve for a simple one. (" Henck's Pocket Book," p. 59.) * \ Let PK = P*K = RaJidPO = P'O' = R' and T O" = T O" = R" and P O T = P f & T = 2 and T O" T' = 2 0'. Assume R' and 0'. Then A = 4 + 20' and O 7 O" : K(y : :sin. WRO" : sin. KO" CX, (R'-R)sin. y 2 A 21 Problem 12. To locate the second branch of a compound curve from a station on the first branch. Let A B be the first branch of a compound curve and D its deflection angle, and let it be required to locate the second branch A B', whose deflection angle is D', from some station B on A B. Let n = the number of stations from A to B, and n' = number of stations from A to B'. Let F = A B B'\ (BAT = nD) n'(nD + n' DQ n + n'. (See " Henck's Pocket Book," p. 61.) Problem 13. To locate a tangent from an inaccessible point on a curve. Let C be the inaccessible point. Run the line to a point B. B E C = 90 C O B. Problem 14. To pass an obstacle on a curve. V 1st method. A C = 2 Rsin. % A O C. 2d " AP= Rtan.y 2 AOC. A O C must be assumed at such a value as, it is supposed, will .carry the line beyond the obstacle. Problem 15. To pass an obstacle on a tangent. (" Mifflin on Rail- way Curves," Prob. 17.) Problem 16. To find the distance across a river, in a prelimi- nary survey. (Communicated by Mr. D. McN. Stauffer.) Jl I j i ' I From A put in the plug B on the opposite side in the line which is being run. Then turn off one degree and put in the plug C. Measure the distance B C ; then AB 100 * or = 57.8 BO. 1.75 Problem 17. To find the radius of a circular arc which shall successively touch three straight lines B D, D E and E C. (From " Rankine's Civil Engineering.") Radius = DE tan. %D + tan. y z E T Problem 18. To connect two tangents with a curve of a given 23 radius when the point of intersection is inaccessible. " Rankine's Civil Engineering.") (From DAE=ADE+AED 180. A D D B = R cotan. sin. AE D E sin: DAE DAE AE = D E sin. AD E sin. DAE. AD;EC=Rcotan. DAE AE. The transit points (marked Tr. P.] are called " plugs," and consist of stakes driven flush with the ground. They are guarded by a stake set on one side with the number turned toward the plug, and under it written (say) "3' off." All the other stakes should have their numbers turned toward the beginning of the line. The following is the form of field-book : (From Shunk.) 1 g I | | o f f o' P | i Q I : f r * f r 23 o 100 N.2000 / W. K2005'W. . 24 50 O* ^ Q Ct- P C + 50 O 50 io6' roo' 25 100 2OO / 300' 2b 100 200' 500' (-IS" ^ 27 100 200' 700' ^c* ^ 28 O 100 200 / 1800' 1400' N 3407'W ^" p + 29 100 200' 1800' f-irfkCj . The final location having been made, the line is divided up into lengths of about one mile each, called sections, and a board placed on end at the dividing station, with the numbers of the sections on the sides. An estimate is 24 made of the amount of earth-work and masonry on each section, and the road is advertised for contract. The con- tractors are each furnished with a printed copy of the quantities in each section, and allowed to take such notes as they require from the map and profile, and walk over the ground, the section boards guiding them to the differ- ent work. CONSTRUCTION. The road is divided up into lengths of about thirty miles, each of which is placed in charge of a " Principal Assistant Engineer." Each of these divisions is sub- divided into lengths of about seven miles, and given in charge of an Assistant Engineer, whose party consists of a Rodman, Chainiran and Axeman. The first work of the Assistant should be to retrace the line and test the bench marks. All plugs should be guarded, and a bench should be made at every culvert. There are various modes of guarding plugs by intersecting lines, by distances from other plugs, or a combination of the two. The best method, where the ground admits of it, is by intersecting lines. A part of a house, or the corner of a window or chim- 25 ney, may often be substituted for one of the above stakes, for a foresight. The note book may be kept in the following form : I \ / 4 \ / H 1 V* V V ! A ! -I*/ EH $> / \* .** L' 6.5 fi vi ^ s t i i i i 'S ' 3 O2 ^ Total Angle pi ^ i 1-1 :: : : Reading jg ^ g 00 I> CO Q1 iH O Deflection S 8 8 8 8 IH H rt ^ b Distance t^ + $* + GO 00 Plu EH f^' U K PH' PH' Station t> CD IO ^ CO OI iH 00 00 00 00 CO GO CO The advantage of this form of field book is, that having 26 first made and checked it in the office, there will be no more calculating of curves in the field, and so much less risk of error. It would always be necessary to run from the same end of the curve, and to use the same transit points ; but this is no objection, as the plugs are all guarded, and it is as easy to set over one as another. Guarding all the plugs saves a great deal of trouble in re- running the line after grading, when it never measures the same as before, and it is difficult to run the old line with- out the same points to run from. At the end of the transit note-book, a page should be devoted to each culvert, giving its station and a little plot of the stakes set, of course drawn roughly ; and a little drawing of the culvert ; also the level of the bridge seat and foundations, etc. The staking out for excavation is done in several ways. In a tolerably flat or undulating country, it is generally done with the level ; on steep hillsides, two rods are used, one ten feet long, and the other of any convenient length, divided into feet by different colors. That which is ten feet long is held horizontally by means of a hand-level laid upon it, with one end resting upon the ground, and the other against the shorter rod, which is held vertically. It is raised until it is horizontal, and the height read off the vertical rod by the Rodman, and noted by the Assistant on a piece of loose paper. He calculates where the slope runs out, and, having checked it on the ground, or made a closer approximation, enters it in a special field-book. The same principle governs the setting of the slope-stakes with the level and level-rod. As a large part of the Assistant En- gineer's work consists in setting slope-stakes, a more minute description is perhaps necessary. They are set opposite the centre line stakes, at the tops of the cuts and bottoms '(, of the fills. If the slope is ^ to 1 and the half width of the road-bed is called b, the horizontal distance of the 27 slope-stake from the centre line is called x, and the height of the slope-stake above the sub-grade is called h ; then By assuming a value of x, measuring it out, and finding the corresponding value of A, with the level or the two rods, the values are substituted in the above equation ; if both sides are the same, the assumed value of x was correct, and the stake should be driven in. If, however, the left- hand side is greater or less than the right-hand, the posi- tion of the stake should be moved toward or away from the centre line an estimated amount, and the process re- peated of taking a new height with the level, and making a new calculation, until the two sides agree. After a little practice, it will not, usually, be necessary to make more than two trials. The following is the form of field book used : The " L. D." signifies the distance out and the height of the ground where the cut or fill strikes the surface of the ground on the left-hand side, while " L. C." means the cut- ting or filling at the half width of the road-bed on the same side. Some engineers look upon the calculation of the point where the slope runs out, in the field, as a waste of time ; and only take the transverse slope, being sure to take it far enough out. They then plot the cross-section, and take the distance to the slope-stake from the plot with 28 a scale. They claim that this method is advantageous, too, becaus* they always run out further than necessary for the slope, and if, afterward, as often happens, the slope will not stand, but slides out a " slip " they still have a record of the amount which slides by measuring to the top of the slide, while, too often, when such an accident occurs, the Assistant finds that he has no note of the slope of the ground beyond his stake, which has been carried away. When such an event occurs, it is better not to slope the cut further up, but to take away the earth at the level of the road-bed for some distance in, to catch any further slide before reaching the track, although the slope may be steeper than was intended. In staking out with the level, it is well to have a number of sheets of paper, fastened together at the edges, for making trial calculations on ; when one is covered with figures, it can be torn off and thrown away, exposing another. The cross-sections should be plotted in a perma- nent record book, to be kept in the office. The area of each should be calculated. For applying the prismoidal formula for calculating the cubic contents, it is requisite to know the middle cross-section between each two that are measured on the ground. The closest approximation to this is the following : Each cross-section is supposed to be transformed into another of equal area, but with a horizontal ground surface, and the depth at the centre of this new cross-section calculated. The depth of the middle section required, is supposed to be equal to the mean of the two end " equivalent centre depths." From this depth the area of the middle section is obtained and substituted in the formula : 5 = -g(A + 4 M + A'), where A. and A.' are end areas, and 3f is the middle area, and I is the distance of end stations apart. Tables have 29 been constructed of "equivalent centre depths" for various areas, and other tables give the cubic contents at once, for a given length and given slopes, from the equivalent cen- tre depths of the end sections. At the top of cuts it is well to have a ditch made on the up-hill side to keep the slope from being washed down. Proper dimensions are : It should be placed about three feet from the edge of the slope. Estimates of the work done are taken up each month. It is important that all papers containing notes of the measurements should be preserved. Although these esti- mates are only intended as rough approximations, the measurements taken will often prove of service in follow- ing estimates. CULVERTS. For finding the proper water-way fo give to culverts, the drainage area of the stream should be discovered if possible. Where county maps are obtainable, this can easily be measured from them. If the drainage area is small, it may often be estimated by walking round it. The water-way may then be calculated by the following formula of Mr. E. T. D. Myers : in which A is the area of the opening of the culvert in square feet, M is the drainage area in acres, and c is a variable co-efficient, depending on the country, and for which Mr. Myers recommends liV in hilly, compact ground, and 1 in comparatively flat ground. In mountainous, rocky country, this value may often be raised to 4. Inquiry should be made of the neighboring people to learn the greatest height of floods in the stream, and the vertical dimension of the water-way may be made equal to the flood height of the stream at the spot. BOX CULVERTS. Rule for laying out on the ground : Take the height of the top of the parapet from the height of the embankment at the centre, and with the remainder (considered as height of embankment) find the side distances with the level as in setting slope-stakes ; then add 18 inches at each end, and if the height of the embankment exceeds 10 feet, add one inch on each end for every foot in height above the parapet. The covering flags are one foot thick and the parapet one foot high, making two feet from top of abutments to top of parapet. For the thickness of abutments, take $ the height of embankment on top of abutments, observing, however, that the abutments must never be less than two feet nor more than four feet thick. To deter- mine the length of the wings, add the height of the open- ing to the thickness of the flags ; one and a half times this sum, added to two feet, will give the distance from the end of opening to end of wing ; the wing to be at right- angles to the drain, unless the latter be askew ; then the wings to be parallel to the direction of the railroad. In- stead of digging deep foundations, the method now em- ployed is to put in a paving made of stones a foot deep, set up on edge, with a curb two feet deep at each face of the drain, and to start the walls on this paving. Should the fall of the drain not exceed 9. inches, make the pavement level, dropping the upper end 9 inches below the surface. Should the fall be greater, make a suffi- 31 cient number of drops of 9 inches each in the length of the drain. At every drop place a cross-sill 2 feet deep ; the wings and parapet to be of the same thickness "as the abutments. The above rule was adopted on the construc- tion of the Junction Railroad of Philadelphia. Box culverts are not usually made of a greater span than three feet. If more water-way is required, two openings are placed, each three feet wide, with a Avail separating them, two feet thick. OPEN CULVERTS. These are generally made of two feet span, with walls two feet thick, with a depth of not more than three feet, founded on a paving, one foot thick. CATTLE GUARDS. These are often placed on each side of a public road crossing, when this takes place at grade. They are built like open culverts with spans varying from three to five feet, and about three feet deep. Stringers 12" x 12" placed 3 feet 11 inches in the clear, support the rails, of a suffi- cient length to rest 5 feet on the solid wall and ground on each side. Two struts, 5" X 12", and 4 feet G inches long, and mortised into each stringer 3 inches, are placed about six inches further apart than the span of the opening, and the stringers are held to them by a rod one inch diameter, 6 feet 5 inches long, with square nuts and long flat washers, placed by each strut. OPEN PASSAGE-WAYS. These are made either with wing-walls or " J " abut- ments. When with wing-walls, the thickness at the base should be calculated like a retaining wall (f the height). The -wing-walls are usually placed at an angle of 45 with the centre line, that angle requiring least ma- sonry. The coping then slopes down at the rate of 2.12 to 1, which is a good proportion for steps, if they are pre- ferred. If the road is for a single track, the " T " abut- ment will be found more economical. The length of the " f " should be so calculated that the earth sloping down it at the rate of 1 to 1, and striking the back of the bridge-seat, and then one end, should just strike the ground at the corner of the bridge-seat, or as near it as the Engi- neer desires. For instance, suppose the distance from A. to sub-grade is 12 feet ; then 12 X IK = 5 + 8 K + x, or x = 10^ = length of B C. JL In staking out for a passage-way, always make the pit a foot larger all round than the foundation is intended to be, so that the quality of the masonry. can be seen. The mason would prefer to fill up the entire pit. If the passage-way is on a curve, having decided where one face should come, turn off from the nearest plug the angle corresponding to the sub-chord at the face, and put in a plug ; set up over this, and turn off the sub-chord to the face of the other abutment. Turn off right-angles from this last sub-chord at each of these plugs, and put in others outside of the 33 pits for the mason to stretch his line by, for the faces of the abutments. Plugs should also be put on both sides of the last sub-chord produced, beyond the pits, to give the centre line of the bridge. This finishes the instrumental work, the other stakes being put in with a tape. A con- venient way of doing this is as follows : t* ] c h 8> 11 JZ Bl ?'* T 1 I* Let A. F be the centre line, marked with plugs at A. and C, and let CD be the face of the neat work. A stake is to be put in at the corner of the pit E, the pit being supposed to be 3 feet larger all round than the neat work. Lining by eye, put in a stake IB 3 feet from C. Then, with the ring of the tape at .Z?-and the 17-feet mark at D, take hold of the 14-feet mark and draw the tape tight ; the 11-feet mark will give the point E. It is well to give the mason a sketch on a piece of paper, giving all dimensions, drawn on the spot by eye without scale, and let him do his own marking out on the founda- tion. The pit is a sufficient guide for putting in the foun- dation. After setting the stakes for the pit, take levels on each one and note in the book ; also note the depth of the pit before the masonry is begun, so that the cubic contents can be calculated. A level has also to be taken at the face before laying out the neat work, to give the height of the 34 neat work to bridge-seat and for calculating the batter and span at the bottom. A 12-feet span bridge, 12 feet high, with a batter of one-half an inch to the foot, would be only 11 feet span at the base of the neat work. STOXE ARCHES. Rankine's rule for the depth of the keystone in feet : For a single arch, D 4/.13 R, For an arch in a series, D = <\/.n R, in which J2 is the radius at the crown in feet. This is for circular or segmental arches. For elliptical arches, for JK substitute ?- when the earth is dry, or - "- when the earth is wet, a being the half-span, and b being the rise. Trautwine's rule is : D rr + 1 A + >2 foot> in which H is the radius of the circle which will touch the crown 3 \ a / a 3 \ Then ? = A C = -% \l + -j^j-l and ^ = B D = g- ^1 + &3 - ) From these equations we obtain the centres C and D. About D with a radius D E = ^ a describe a circular arc, and about (7 with a radius C E = a ? describe another arc ; the intersection of these at E will be the third centre. When brick is plentiful, circular culverts are often em- ployed. They require no foundation except for the face walls. For the thickness, one brick (nine inches), is suffi- cient for a span less than six feet. Add one more brick for each six feet more of span up to thirty feet. The following are two cheap forms of centring : 14-/ee< span. Frames %% feet apart. 3Y 24-/ee span, 8-feet rise. Frames B% feet apart. Post mortised (by slight tenon) into chord. Arch pieces pinned together and halved in chord and post ; braces spiked at ends ; and at intersection with post a i-inch bolt is used. Centres should be removed from arches, unless laid in a very quick-setting mortar, within a few days after their completion. In stone arches the parapet should not be made too high, or it may be pushed over by the bank ; it is well to proportion it like a retaining wall if more than one or two courses high. Some loose stones laid flatwise behind it will relieve the thrust of the earth. In designing centres, allow ^OT of the span for settling of the arch, unless built very slowly and with great care. RETAINING WALLS. Let 6 = the breadth at the bottom = 1. h = the height. t = the thickness at the bottom. w = the weight of a unit of volume of masonry. w' = the weight of a unit of volume of earth. = the slope of the bank above the wall. q> = the angle of repose of earth. 41 21 Let j = the inclination of the foundation pit to the horizon. " q = a constant of safety distance from middle of base to point where the line of resistance cuts base It is always between and *. Letg' distance from the middle point of base to point where base is cut by a "rertical line through the centre of gravity thickness at base. total weight of masonry Let n = - -~ It is always less than . .tal cos. 4/cos. 8 6 cos. 2 a> Let w^ = w cos. - cos. 6 + /cos. 8 cos. 8 = 35 ; tt> t = .27 w/. n = y 2 and./ = in addition, * - - \/ - 27M? t h Z(qq r ) w. If we suppose the wall to be just stable without the least excess of strength, q = i It is customary, however, to give q a smaller value, so that there will be an excess of strength ; otherwise the pressure would be concentrated at a point, and it would split off or crush. The English engineers make q = .375. For q' we can assume the following values for walls of different heights : Height of wall. 20 35 55 100 .11 .14 .14 .16 w' 100 For first-class masonry we may take = w 165 100 For dry sand-stone rubble we may take " = 120 t Then if the wall is over 100 feet high, of first-class masonry, = .51 h dry rubble ' ' .60 " " between 55 & 100 ft. high, first-class ' .49 " dry rubble ' .58 If the wall is between 35 & 55 ft. high, of first-class .48 dry rubble ' .56 " " less than 20 " first-class ' '. .45 dry rubble .53 When* the wall is rectangular in section : t If of first-class masonry, = .27 h " dry rubble " " .32 When = This is a combination of the king and queen post systems. To prevent the point B from rising on the application of a weight at Z>, braces are often introduced at A E and C F. The force W, acting upward at JB, produces a force equal to Wsin. a in the direction of C F ( C F being at right angles to F B] and acting at B, which has a tend- ency to break the beam transversely at F. The formula 48 for the breaking -weight of a beam fixed at one end and & d 3 loaded at the other is W= S The pressure on C F L is also TFsin. a, which produces a transverse strain in A B at C) which can be calculated in like manner. The trussed girders of the following forms may be looked upon as king and queen post trusses inverted : They are objectionable as being a combination of two systems, and it is impossible to tell just how much of the load will be borne by the beam acting as such, or by the rods transmitting a portion of the strain to compress the beam. It depends on the adjustment of the rods. D F The "Warren girder," or "triangular truss:" Let W = the weight on one panel =the weight on FC due to uni- form load or weight of bridge. pj7/ _ the weight on one panel = the weight on .FCdue to the variable load. " n = the number of panels A E, E B, etc. If the load is on the upper chord, struts FJB, Jf I, etc., are introduced, and if on the lower chord, rods D U, C H, etc., so that each apex of the triangles will bear a load. The maximum strain on one of tnese struts or ties would be w+ w. 49 n _ i Ilie maximum compression on A D = ^ - (W + W) sec. 0. 1 V 2 /" Hie maximum compression on D B = -g - - W sec. 0. f Q&, ZAh^frt asimum tension on D B The maximum compression on B O = j (m - 3)(n - 8) - (2x 8)||^ + (-3,(n-2) ^^ The maximum tension on 1? <7 = 2 ^ TF' sec. 9. (^-M- & The maximum compression on C I = >> - ~ TF' sec. 9. f f i< y The maximum tension on C J = { ( .-4K.-)-(xo} The maximum compression on IL The maximum tension on IL = - 2 n W sec. 6, /V-^C < etc., etc., to the middle of the bridge, when the order will be reversed. The strains on the chords are greatest when the bridge is uniformly loaded. We then have : Tension on A B = -= (W + W) tan. Q. n _ i Tension on B I - g (W + W) tan. 6 , ) (W+ W')tan.Q. + (n - 2)(n - l)-2 ^ - ( W + W) tan. Q 2n. Tension on I M = tension on B 1+ H (n 4) (n 3) (3 X 4) / } -, tw + W] tan. + j (n - 5) (n - 4) - (4 X 5) | J ^ - etc., etc. Compression on DC [n l -2- (W + Compression on C L = compression on D C + [ | (n 3) (n 2) (2 X 3) + (B _ 4 , __, x ! etc., etc. If the roadway is on the upper chord, the in- clined piece, A D, in the last figure, is sometimes left out, and the bridge built as follows : IV\/\/\/ x o xi d A D, J) JB, D E, etc., sustain the same amount of strain as before, but what was then compression is now tension, and vice versa. A B has the same amount of strain as A B in the other figure, and D C as D C, the kinds being reversed as before. The part of the lower chord, D O, might, in this case, be dispensed with, were it not that it is of use in the lateral bracing, and must, therefore, be introduced. The triangular truss is often built with two or more sets of different triangles forming the bracing ; when of more than two, it is called a lattice truss. The strains on these can easily be represented in formulas as above, but are omitted for the sake of brevity. Each separate system of 51 triangles has strains like the previous form, but inde- pendent of each other. T G The Howe truss : Let n = the number of panels. " W = the weight on one panel due to the uniform load. " W = the weight on one panel due to the variable load. n ^ (W+Wf) ^. Maximum compression on FB Maximum compression on G C -3 (n-2)(n-l) w ) FB_ W + Maximum compression on H D _ j n-5 . (n-8)(n 2) Maximum compression on 1 E ( Ti 7 W FB ~BG~ 'FB BG, etc., etc. By continuing this process past the middle of the bridge, we obtain the maximum strains on the counter- braces, and when a strain becomes minus, it shows that be- yond that point the counter-braces may be left out. Maximum strain on B C = ^ (where n = the number cf the frog). g cos. Y z a ~ sin.y^F. The value of the radius of the turnout differs in the two cases as below. _ff Eadiu, of t^out = To find the distance from the point of switch to point of frog on a tangent : 1 + cos. F _ g x ~ 9 sin. F ~ tan. %F. It is usual to designate frogs by numbers which ex- press the relation between the base and altitude of the tri- angle forming the point of the frog. Thus, a No. 8 frog is one whose length is 8 times the base; D C = 8 A B. The above equation then reduces to x = 2 g n, where n = the number of the frog. The radius of the turnout is = . = -t q. sin.F For a No. 8 frog, FV 9^'. It is the usual practice to spike down 5 feet on each side of the frcg straight, or, calling the distance from point to end of frog two feet, there would be 12 feet straight. The g in the formula (for a 4 feet 9 inches gauge) would be reduced to 3 feet 10^- inches. In this case, then, x = 62 feet and R 495.72, or about an 11 35' curve. Five feet of the switch rail is spiked fast. In order to have a throw of .5- inches, the switch rail should be 27 feet long. The distance from the movable end of the switch rail, or point of switch, to the point of the frog is, then, 47 feet. This is an ordinary switch on railroads. To find the distance from frog to frog on a crossing : Call the distance from point of frog to point of tangent = c. The distance measured on the track from frog to frog : Then, d = \ e a + a 2 For a No. 8 frog, when g 4f , a = 7^-, and c = 7, e = 12.67, d = 14.60. To find the proper angle for the middle frog in a three- throw switch : Ver. sin. % F' = sin. 2 % F. When two No. 10 frogs are used, we find F' = 8 6', and the third frog should be a No. 7. When two No. 8 frogs are used, the other should be a No. 5.64, or, say, 6. Bill of crossing lumber, single switch : 3 ties 9% feet long. ...10 2 I 1 switch tie 15 feet long, 16 inches wide. 70 Bill of crossing lumber, double connection : 6 ties 9* feet long. 2 10 4 10 4 11 4 11 2 12 16 19 2 switch ties ... .15 feet long, 16 inches wide. Bill of crossing lumber, triple connection : 2 ties 9 feet long. 3 10 2 11 2 12 8 13 1 14 2 15 1 16 3 17 3 18 4 19 2 20 1 switch tie 15 feet long, 16 inches wide. CROSS TIES. These should be hewed (not sawed or split) on two sides, cut square at the ends, and stripped of the bark before de- livery. They should be 8 feet long and 8 inches thick. Three-fourths of them should measure not less than 8 inches across the hewed surfaces, and one- fourth not less than 10 inches. They should be piled in square piles of about 50 each, the ties crossing each other at right angles in alternate layers. Each pile should be separated from the rest, so that a man can pass around each one to inspect the ties. Public road crossings at grade : The space between the tracks is covered with plank, 3f 71 X 8 inches, 16 feet long, spiked to the ties, and leaving 4 inches clear by the rail for the wheel flanges. Planks are also spiked to the ties on the outside of each rail. RAILS. A width of 4 inches is sufficient to prevent the rails from cutting into oak ties, and 4$ inches for chestnut ties, when not spaced more than 2^- feet apart. If the base is made more than this, the difficulty of bending the rails to a curve becomes an objection. The stem of an iron rail need not be more than one-half an inch thick, nor that of a steel rail irore than & of an inch. 60-pound rails are made 4% inches Zf l L high, 50-pound 4 inches, and those under 50 pounds 3 inches. (See Engineering, Vol. 18, p. 369.) WATER STATIONS. Passenger engines on the Middle Division of the Penn- sylvania Railroad, where the grades are very light, run at a rate of 35 miles per hour with seven cars ; and, when mak- ing frequent stops, one tank of water, containing 2,400 gallons, lasts for 2% hours, or 78 miles. The engines, how- ever, take in water, actually, every 45 miles. A freight train on the same division, with a full tank, can run at a speed of 14-|- miles an hour for 2 hours 50 minutes, or 41-^ miles with one tank of water. As, however, they have to stop at shorter intervals to allow passenger trains to pass, or to pass each other, they utilize the time of waiting in fill- ing their tanks. On the Mountain Division of the same railroad, freight trains with a full load on an almost con- tinuous grade of 1& per cent., use a tank full of water, con- taining 3,000 gallons, in 1 hour 15 minutes, or in going 15 miles. It is thus seen that 15 miles is the extreme distance apart for water stations with grades of two feet in a him- 72 dred, while 40 miles would do with very light grades. It would be well, however, if water is plenty, to have them every 10 miles, or of tener. In a hilly country, streams can generally be dammed up, which will give a gravity supply. The outer slope of the dam may be built of stone, like a retaining wall, or may be of earth at its natural slope. The inner slope should be at the natural slope of clay in water, which is 3 to 1. There should be a layer of good clay on the inside, two feet thick. Cast-iron pipe, 6 inches in diameter, is used to convey the water from the dam to the track. This should have a fall of at least 1 or l feet per 100 for 1,000 feet from the dam. For the remaining distance, if the water is brought so far, the fall had better be not less than .333 per 100. If, how- ever, it is impossible to give it a continuous down grade, it may be laid undulating, so long as no portion rises above the " hydraulic grade line," as defined in " Trautwine's Pocket Book." If it is so laid undulating, it will be neces- sary to place an air cock at every " summit," and a mud cock (blow-off cock) at every " valley." At the track, a " stand pipe " or " plug " is placed, which rises to an outlet 9 feet above the rail. A valve controls the outlet, within reach of the engine-driver. A piece of rubber hose, 7 inches in diameter, 10 feet long, is fastened on the end of the pipe, to insert into the opening in the tender. It may often happen that a dam cannot be made, or there is not enough water in the stream to furnish a continuous supply. A tank is then placed at the side of the railroad. This is a tub made of white pine, 18 feet in diameter at the bottom, and 1 7 feet at the top, 8 feet deep. The bottom is 3 inches thick , and the staves 2^ inches thick. There are 6 iron hoops, X 3 inches ; two placed close together at the base, and the others at intervals increasing toward the top. The bottom is let into a groove in the staves, but the ends of the staves are let into the floor, so that the bottom bears 73 over its whole surf ace on the floor. The tub is supported on three trestles of 10 X 10 inches stuff, placed 6 feet 6 inches apart, on walls 18 inches thick, built parallel to the track, and finished off one foot above the rail. On these trestles, joists 4 x 12 inches are placed one foot apart, which support the floor of two-inch stuff, on which the bottom of the tub directly rests. Where a greater supply is required, or a more permanent structure, and an adjacent hill permits it, stone reservoirs are made 40 feet in diameter and 8 feet deep. They are built below the surface of the ground. The walls are built of common mortar, with a lining of brick well wet and thoroughly beddt d in cement. The bottom is covered with a layer of stones about the size of a walnut, 4 inches deep, and made into a concrete with cement; and when it is set, another layer of the same thick- ness is put in. These reservoirs are covered with a house. Where a gravity supply cannot be obtained, water must be pumped into the reservoir with a steam engine or windmill. Six-inch pipe, of a thickness of ^V of an inch, is made to lie in lengths of 12 feet. Each joint requires 8 pounds of lead and a quarter of a pound of " gasket," or loosely twisted rope, which comes for the purpose. COALING STATIONS. A freight engine with its load, on very light grades, con- sumes about 160 bushels of bituminous coal in going 131 miles. This will give some basis for calculating the dis- tances at which coaling stations must be provided. PASSENGER STATIONS. Description of Cresson Passenger Station, Pennsylvania Railroad : One story high, 70 X 40 X 12 feet high, with sloping 74 roof. Posts or " studs " are sot 6 X 7 inches at the corners, and 5X6 inches at points 5^ feet apart. These are braced by horizontal pieces, 3X4 inches, placed about 4 feet apart, except at the windows and doors. Diagonal braces, 4X6 inches, are placed at the upper corners, framed into the posts and a beam, 6X7 inches, which f onus the tie beam of the roof truss. The latter is a king post of 6 X 6 inch pieces, with secondary king-post . trusses, abutting toward the centre against a straining beam, all 4 X 6 inches. The ridge pole is 2 X 10 inches, with purlins 4X7, and rafters 3X5 inches, spaced two feet apart. Joists, 3 X 10, spaced 1^ feet apart. Flooring, l inches, worked. Roof sheeting, 1 inch. Platform flooring, 2 inches. Platform joists, 3X9^ inches. Weather boarding, of an inch thick 9 inches wide, with % of an inch stripping, 2 inches wide. Partition of $ of an inch stuff. Plastering lath, 3 feet long. Water station at Gallitzin, Pennsylvania Railroad: " Balloon frame," 22 feet 8 inches by 22 feet 8 inches by 18 feet 6 inches high. Wall plates, 3X8 inches. End posts, 4 X 4 inches. Studs, 2X4 inches, 18 inches apart. Diagonal pieces, 1^ X 3 inches, 2 feet 6 inches apart, meas- ured vertically. Rafters, 2X6 inches. Ridge pole, 2X8 inches. Joists, 4 X 12 inches. Siding of $ of an inch worked boards, tongued and grooved. Sheeting, ditto. Slate roof. It may here be remarked that when a plank is nailed to a post or joist, or other wooden substance, a nail is used, of such a length that it will go twice as far into the post or joist as the thickness of the plank. Thus, for f-inch stuff, use 2^ inches longer 8 penny nails, for 1-inch stuff use 3 inches long or 10 penny nails. TELEGRAPH LINE. The number of poles to the mile varies from 26 to 42. 75 The size of wire varies from 320 to 380 pounds per mile. The poles should not be less than 5 inches in diameter at the top, nor less than 25 feet long. If green, they should be charred 5 feet from the bottom. If any are split at the lower end, the parts should be nailed together before putting in the ground ; otherwise, the spring of the wood will pre- vent the earth from packing around them. The cross-arms are made of 3 X 4 inches, 3 feet long, white pine, painted white, one bolt for each cross-arm, inch diameter, 8 inches long, square head and nuts, and wrought washers. Sixty miles of line will require at each end a battery of 15 cups (Grove's). These cups require to be re-charged twice a week. A battery of 30 cups requires 1 carboy or 200 pounds of nitric acid, 25 pounds of sulphuric acid, and 1 pound of zinc per cup, every month. 1 1 t| "** ^^f ^V S3 u. JU %BDWSO^ ^W( .^EUNIVERSyA ,vvld ^v'_ ^^^ gxH i MIBRARY0/- ^UIBRARY ' ft ' 0073396905 A\\E UNIVERJ//,