^. 
 
 
 IMAGE EVALUATION 
 TEST TARGET (MT-S) 
 
 ^. 
 
 4 
 
 
 ./ 
 
 ^< 
 
 ^ 
 
 ^ 
 
 « 
 
 1.0 
 
 1.1 
 
 25 
 2.2 
 
 ■a 12.8 
 
 US Itt 
 
 u HA 
 
 Sf U£ 12.0 
 
 
 lai 111^ IJ-^ 
 
 I 
 
 FhotograiJiic 
 
 ^Sciences 
 
 Corporation 
 
 r<N- 
 
 s> 
 
 23 WIST MAIN STREET 
 
 WE»STER,N.Y. 145M 
 
 ( 71* ) •72-4503 
 
 4R> 
 
 i 
 

 CIHM/ICMH 
 
 Microfiche 
 
 Series. 
 
 CIHIM/iCIVIH 
 Collection de 
 microfiches. 
 
 Canadian Institute for Historical Microraproductions / Institut Canadian da microraproductions historiquas 
 
 /• 
 
Tachnieal and Bibliographic NotM/Not«s taoliniquM at iiibliographiquat 
 
 Th« 
 tot 
 
 Tha Inatituta liaa attamptad to obtain tha baat 
 original copy availabia for filming. Faaturaa of thia 
 copy wlilch may ba bibliographlcally unlqua, 
 which may altar any of tha imagaa in tha 
 raproductlon, or which may aignificantly changa 
 tha uaual mathod of filming, ara chacicad balow. 
 
 D 
 
 D 
 
 D 
 D 
 
 D 
 
 D 
 
 Colourad covara/ 
 Couvartura da coulaur 
 
 r~1 Covara damagad/ 
 
 Couvartura andommagAa 
 
 Covara raatorad and/or laminatad/ 
 Couvartura raataurAa at/ou pallicuMa 
 
 I I Covar titia miaaing/ 
 
 La titra da couvartura manqua 
 
 Colourad mapa/ 
 
 Cartaa gAographiquaa an coulaur 
 
 Colourad ink (i.a. othar than blua or black)/ 
 Encra da coulaur (i.a. autra qua blaua ou noira) 
 
 r~n Colourad plataa and/or illuatrationa/ 
 
 Planchaa at/ou illuatrationa an coulaur 
 
 Bound with othar matarlal/ 
 Raiii avac d'autraa documanta 
 
 Tight binding may csiuaa ahadowa or diatortlon 
 along intarior margin/ 
 
 La re iiura sarrte paut cauaar da I'ombra ou da la 
 diatortlon la long da la marga intiriaura 
 
 Blank laavaa addad during ras'ioratlon may 
 appaar within tha taxt. Whanavar poaaibla. thaaa 
 hava baan omittad from filming/ 
 II aa paut qua cartainaa pagaa blanchaa ajouttoa 
 lora d'una raatauration apparalaaant dana la taxta, 
 mala, loraqua cala Atalt poaaibla, caa pagaa n'ont 
 paa 4tA fiimtea. 
 
 Additional commanta:/ 
 Commantairaa aupplAmantairaa: 
 
 L'Inatitut a microfi.-ni la malllaur axamplaira 
 qu'ii lui a At* poaaibla 69 aa procurar. Laa dAtaiia 
 da cat axamplaira qui aont paut<Atra uniquaa du 
 point da vua bibllographiqua, qui pauvant modlfiar 
 una imaga raproduita, ou qui pauvant axigar una 
 nvodification dana la m^thoda normala da filmaga 
 aont indiquAa ci«daaaoua. 
 
 Th< 
 
 D 
 D 
 D 
 [2 
 D 
 
 D 
 D 
 
 n 
 
 Colourad pagaa/ 
 Pagaa da coulaur 
 
 Pagaa damagad/ 
 Pagaa andommagAaa 
 
 Pagaa raatorad and/or laminatad/ 
 Pagaa raatauriaa at/ou palilculAaa 
 
 Pagaa diacolourad, atainad or foxad/ 
 Pagaa dteolortea, tachattoa ou piqutea 
 
 Pagaa datachad/ 
 Pagaa d^tachAaa 
 
 Showthrough/ 
 Tranaparanca 
 
 Quality of print variaa/ 
 Quality inAgala da I'impraaaion 
 
 Includaa aupplamantary matarlal/ 
 Comprand du matirial auppMmantaIra 
 
 Only adMon availabia/ 
 Saula Mitlon diaponibia 
 
 Pagaa wholly or partially obacurad by arrata 
 allpa, tiaauaa, ate, hava baan rafilmad to 
 anaura tha baat poaaibla imaga/ 
 Laa pagaa totalamant ou partiallamant 
 obacurciaa par un fauillat d'arrata, una palura, 
 ate, ont At* fllmAaa A nouvaau da fa9on A 
 obtanir la malllaura imaga poaaibla. 
 
 oft 
 filn 
 
 Ori 
 baf 
 tha 
 atoi 
 oth 
 fira 
 aioi 
 ori 
 
 TN 
 aha 
 
 TIN 
 
 Ma 
 diff 
 ant 
 bag 
 
 rigl 
 raqi 
 mat 
 
 Thia Itam la filmad at tha reduction ratio chackad balow/ 
 
 Ca doeumant aat filmA au taux da rAductlon indiquA ci-daaaoua 
 
 10X 14X . 18X 22X 
 
 
 
 
 28X 
 
 
 
 
 30X 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 y 
 
 
 
 
 
 
 
 
 
 
 
 
 n 
 
 
 12X 
 
 16X 
 
 
 
 
 aox 
 
 
 
 
 24X 
 
 
 
 
 2SX 
 
 
 
 
 32X 
 
I 
 
 itails 
 idu 
 odiffiar 
 'un« 
 
 Th« copy filmed hw hat bean raiiroduead thanka 
 to tha ganaroalty of: 
 
 Broek Unhrtnity 
 StCathiriiMf 
 
 Tha imagaa appaaring hara ara tiM baat qualltv 
 poaaibia oonaidaring tha condition and laglbiiity 
 of tha original copy and in kaaping with tho 
 filming contraet apaeif ieationa. 
 
 Original eoplaa in printad papar covara ara fllmad 
 beginning with tha front oovar and andlng on 
 tha iaat paga with a printad or illuatratad impraa- 
 aton, or tha back eovar whan appropriate. AS! 
 othar originji copiaa ara fllmad beginning on the 
 firat puge with a printad or illuatrated impree- 
 aion. and ending on the Iaat page with a printed 
 or illuarrated impreealon. 
 
 L'a«amplaira filmi fut reproduit grice i la 
 gAnAroaitA da: 
 
 Broek UnKrartity 
 St CitharinM 
 
 Lea imegea suivantaa ont Ati raproduitea avac la 
 piua grand soin, compta tanu da la condition at 
 da la nettetA de raxampieire film*, at an 
 conf ormM avoc lea conditiona du contrat da 
 fllmage. 
 
 Lea axempleirea originaux dent la couvartura an 
 papier eat imprim^e sent filmto an commandant 
 par la premier plot et en terminent aoit par Ic 
 darnlAre pege qui somporte une amprainta 
 d'Impreaaion ou d'iiiuatration, aoit par la second 
 plot, selon le cea. Toua lea autrea axempleirea 
 originaux sent film4a en commen^ant par la 
 pramlAre pege qui comporte une empreinte 
 d'Impreaaion ou d'illuatration et en terminent per 
 le demlAre page qui comporte une telle 
 empreinte. 
 
 The leet recorded freme on eeeh microfiche 
 ahaM contain the symbol «^(meening "CON- 
 TINUED"), or the symbol ▼ (meening "END"), 
 whichever appliee. 
 
 Un dee aymbolee suiventa appareltra sur la 
 damiAre imege de cheque microfiche, selon le 
 caa: la symbole ^ signifle "A SUiVRE". le 
 symbole y signifle "FIN". 
 
 Maps, platea, charta. etc.. mey be filmed et 
 different reduction retioa. Thoae too ierge to be 
 entirely included in one expoaure ere filmed 
 beginning in the upper left hend comer, left to 
 right and top to bottom, aa many framea aa 
 required. The following diegrema illuatrate the 
 method: 
 
 Lea cartea. planchea, tableaux, etc., pauvent Atra 
 filmAa i dea taux de rMuction diffArenta. 
 Loraque la document eat trop grand pour Atra 
 reproduit en un seul clichA. il est fiimA A partir 
 de Tangle supArieur gauche, de qeuche A droite, 
 et de haut mn bee, en primant le nombre 
 d'Imegee nAceaaaire. Lea diagremr.tea suiventa 
 Uluatrant le mAthode. 
 
 rrata 
 to 
 
 pelure, 
 nA 
 
 □ 
 
 32X 
 
 1 
 
 2 
 
 3 
 
 1 
 
 2 
 
 3 
 
 4 
 
 5 
 
 6 
 
REPORT 
 
 ON riiK 
 
 RENEWAL 
 
 OP 
 
 NIAGARA SUSPENSION BRIDGE. 
 
 BY 
 
 LEFFERT L. BUCK, 
 
 CIVIL ENGINEER. 
 
 xeso. 
 
 NEW YORK: 
 C. W. Ames & Co., Pkintkks, 96 Chamukks Stukict. 
 
 1881. 
 
 145245 
 
Names (if memhcrs of the Boards of Directors of the 
 Niagara Falls Suspension and Niap:ara Falls International 
 riridp^e Companies during- the progress of the work : 
 
 ^iiiei'traii )Bom;b. 
 
 Hon. r,OKK\Z() hi KKOWS. I'iun. 
 
 AI-MKUT S. WAKXKR, Sk( kktvuv. 
 
 Hon. I{. L. BURIKIWS. 
 
 UKU. I,. BIKROWS. 
 
 Col. V. A. ROKBLINfJ. 
 
 A. STKWART. 
 
 C. n. MOORK, 
 
 JOSEPH A. AVOODRUFF, I'kks. 
 J. F. McGLASIIAN, SEc.iKTAiiv. 
 Hon. J. R. BENSON. 
 THOMAS R. MERRITT. 
 Right Rev. T. B. FULl ER 
 RICFIARD MILLER. 
 JOHN L. RANNKY. 
 
 WILLIAM (.. SWAN, Superintauknt. 
 
To the Presidents and Gentlemen of the Hoards 
 of Directors of the Niagara Falls Supension and the 
 Niagara Falls International Bridge Companies: 
 
 Gkntlemen: 
 
 Having completed the work of rein- 
 forcing the Anchorage and renewing the Suspended 
 Superstructure of your bridge, I respectfully submit 
 the following report. 
 
 LFFFFRT L. lUXK. 
 
 December 13th, 1880. 
 
I 
 
REPORT 
 
 ON THE REINFORCEMENT OF THE ANCHORAGE AID RENEWAT, 
 OF THE SUSPENDED SUPERSTRUCTURE OF THE NIAGARA 
 RAILROAD SUSPENSION BRIIKJE. 
 
 From the inception of the project of spanning the chasm of 
 the Niagara River below the Falls, with a siisj)ension bridge 
 for railroad puri)oses, to the year 1855, when the bridge was 
 completed and opened to traffic, it was considered bj all as a 
 bold undertaking and by some engineers, even, as an impracti- 
 cable one. But tlie bridge has been in constant use for twenty- 
 five years and under constantly increjising traffic, thus demon- 
 strating its adaptability to a locality requiring such long spans. 
 
 In spite of its succe>^s, however, it has been an object of 
 constant solicitude to the traveling public. The frightful 
 chasm that it spans would naturally excite the fears of most 
 people, but this has been greatly enhanced by doubts as to tlie 
 condition of the cables and their anchorage. 
 
 The object of this Eeport is to show what the real condition 
 of the cables and anchorage was, and also to indicate what has 
 been done ior their improvement, as well as that of other parts 
 of the bridge. 
 
 In order to explain this clearly, it is necessary to insert here 
 the following general description of the structure as it was. 
 
 DESCRIPTION. 
 
 The bridge consisted of the following members : " 
 
 1st. Two pairs of iron wire cables resting on masonry towers 
 at each end of the bridge. The ends of the cables are secured 
 by means of chains to suitable cast iron anchor plates bedded 
 in the rock forming each bank of the river. Two of these 
 cables have, at mean temperature, a versedsine of 54 feet, and 
 are designated upper cables. The other pair having a versed 
 sine of 64 feet, are called the lower cables. 
 
6 
 
 2(1. Tlie Suspended Siiperatructure. 
 
 Tliis consisted of two floors, placed at a vertical distance 
 apjiit of 17 feet and co!inocted by posts and rods in such a 
 man nor as to form a trussed tube. 
 
 At each five feet in the length of the trusses, two wire rope 
 suspenders connect the up{)er floor with the uj)per cables. In 
 the Slime manner the lower floor is suspended to the lower 
 cables. 
 
 CONSTRUCTION OF THE CABLES. 
 
 Each cable is composed of seven strands or bundles of wire. 
 Each strand is made up of 520 scant No. 9 wires laid })arallel, 
 and at each end formed into a loop which fits into a groove in 
 a U shaped cast iron shoe. The seven strands are bound into 
 one bundle, of 3,640 wires, which is served closely with wire 
 over the whole length, with the exception of about 13 feet, at 
 each end, and of about 10 feet of the portions resting on 
 the towers, thus forming a cylindrical cable \0^ inches in 
 diameter. 
 
 The tops of the towers are each covered with a cast iron 
 plate, 8 feet square, bedded in mortar. The upper surface of 
 this plate is planed to a true surface. On it rest a number of 
 turned iron rollers 5 inches in diameter. On these rollers rest 
 the saddles, consisting of heavy castings, whose undersides are 
 planed. The top of each saddle has a groove of semi-circular 
 section in which the wires of the cable lie, each cable having a 
 separate saddle. The planes of the curves of the cables, 
 between the towers, are inclined in such a manner as to bring 
 those of each pair nearer together at the middle of the span, to 
 give lateral stability to the bridge. From the towers to the 
 anchorage the cables diverge from the center line of the bridge 
 sufRciently to make the platie, containing the portion each 
 side of the tower, vertical. The wire forming the cables was 
 boiled in linseed oil before it was laid, and as the cables 
 were made the interstices at the shoes and towers were flushed 
 with boiled linseed oil and Spanish brown paint. Then the 
 whole length of the cable was flushed with the same as the 
 serving progressed. 
 
 ti 
 
i 
 
 ANCHOR AUK 
 
 Each end of each cable lias a separate ancliornge tlie derfcrip- 
 tion of wliich will be assisted by referring to Plate 1. 
 
 A rectangular pit or shaft, 3 ft. x 7 ft. in plan, was snnk 
 vertically into tlie rock, to a deptii of 25 feet, with the bottom 
 enlarged to form a chamber 7 ft. square. An anchor phite, 
 6 ft. 6 in. stpiare and having seven rectangular opetiings tlirongh 
 it to receive the lower links of the anchor cluiin, is set in the 
 chamber, the links put in position and secured by a 8.1 inch 
 diameter pin passed through their heads and underneath the 
 plate. From the plate the chain passes vertically upward to 
 the surface of the rock. From this point the joints of the 
 chain are at points of a vertical curve of 25 ft. radius. The 
 joint at the upper end of the curve lorming the j)oint of 
 tangency with the line of the cable. Beyond this joint is 
 another length of chain composed of nine links, each bar of 
 which is 10 feet long and 7 in. x If in. section. Four of these 
 links alternate with the shoes of three of the strands of the 
 cable and are secured to them by a 3^ in. diameter pin passing 
 through links and shoes. The remaining five links are in like 
 manner connected with the remaining four shoes of the cable 
 strands, as shown in elevations (Fig.s. 1 and 2). 
 
 The anchor plates are secured in tlie chambers by means of 
 neatly fitted stone blocks set in cement mortar, the whole pit 
 being solidly filled with cement masonry and the interstices 
 around the bars grouted. Above the rock and np to the end 
 of the chain the whole is enclosed in a solid wall of masonry, 
 heavy blocks of which form supports of the joints of the 
 curved portion of the chain. Formerly the strands were also 
 covered with masonry and the whole grouted, the intention 
 being to preserve them from corrosion. 
 
 CONSTRUCTION OF THE OLD SUSPENDED 
 SUPERSTRUCTURE. 
 
 As first built the superstructure was as follows (see right 
 hand half of transverse view and side elevation of old truss, 
 plate IV.) 
 
8 
 
 1st A aeries of transverse bents, one at eacli five feet in tlie 
 length of the bridge, or 161 in nil. Kucli bent consisted of the 
 following parts: 
 
 a. Lower floor beam. This was made of two pieces of 4 in. 
 X 1() in. pine timber, 24 feet long, set on edge, parallel and 
 with a 6 in. space between them. 
 
 b. Two posts. Kach post was composed of two pieces of 
 4 in. X in. pine placed parallel and separated 4^ inches by 
 packing blocks. The lower ends of the posts were secured l)e- 
 tween the two parts of, and (lush with, the lower edge of the 
 lower floor beam. The clear distance between the posts of a 
 bent was 19 feet. 
 
 c. U{)per floor beam. The upper beam was the same as the 
 lower, exc(;pt that it was 26 feet long. The top of the post 
 was flush with the top of the beam and clamped between its 
 two parts. 
 
 d. Knee braces. Each bent ^as stiffened transversely by a 
 4 in. X 6 in. pine knee brace from the under side of the upper 
 beam to the inner face of the post. 
 
 2d. Longitudinal members. 
 
 a. Upper and lower chords. 
 
 The upper chords each consisted of two thicknesses of 
 2 in. X 12 in. white oak plank in lengths of 5 feet, the joints of 
 one thickness being at the middle of the lengths of the other. 
 They were bcHed together with eight ^ in. bolts to each length. 
 These chords laid on tops of posts and beams. 
 
 The lower chords were each made of three thicknesses: one 
 of pine 2^ in. x 12 in., one of pine 4 in. by 12 in., and one of 
 oak 2 in. x 12 in., all in 6 feet lengths and bolted together with 
 eight ^ in. bolts to each panel. The lower chords were laid under- 
 neath the ends of the posts and the undei*sides of the lower 
 floor beams. 
 
 b. Track Stringers. 
 
 There were two deep longitudinal track stringers built 
 partly above and partly below the upper floor beams. The 
 portion above the floor beams was made of 4 in. x 16 in. pine 
 planks piled closely, broadsides horizontal and to a height 
 above the beams of 2 ft. 6in. The rails for the track were 
 
ect in the 
 ted of the 
 
 s of 4 in. 
 allel and 
 
 pieces of 
 iiches by 
 3ured he- 
 re of tlie 
 of a 
 
 3 
 
 JStS 
 
 le as the 
 the post 
 ween its 
 
 e]y by a 
 e upper 
 
 3sses of 
 joints of 
 e other. 
 I length. 
 
 ses: one 
 one of 
 er with 
 i under- 
 e lower 
 
 3 built 
 I. The 
 n. pine 
 height 
 £ were 
 
 
 i 
 
 laid on the tops of thcso shIiilmth. Tlic portion ol cjich 
 stringer below the bcjMiis wiis nuidr <»1 two picfCH of pine 
 tiniber, one piece being 12 in. \ 12 in. and the otlier 12 in. x 
 10 i»i. Thev W(!re \\\m\o eoiitinuoiis by searle*) and keyed 
 spliceH. The spaces between the upper iiiid lower portions ol 
 the stringers and from t)nc beiini to anoliier were tilled by 
 bridging pieces 4 in. tliicU. There were 5 long ])olts (1 itich 
 diameter) to eacii panel lenglli ol stringer extending llirongh 
 from top to i)ottoni. 
 
 c. Ilarnl rails. At each side of the npper lloor was a iieavy. 
 trussed lian<l rail, exteiidin^r iho whole length of the ])ridge. 
 At each beam a \l in. diameter bolt passed Irom the top ol tiie 
 hand rail down between the pjirts of the beam mid ihroujih an 
 8 in. X 9 in loiii^itinlinal pine string piece. This string piece 
 was made continuous ])y splicing. 
 
 d. Upper lloor. Tlu; planking of the upper lloor was laid 
 lengthwise and fastened to tlic beams by means of screws. 
 The |)t)rtion between tlie track stringers was of 4 in. j)lank, 
 the renniinder was 2 in. thick. 
 
 c. Lower lloor planking. The first course of ])1atdc was of 
 pine 2 in. thick. The top course of oak 2 in, thick. Both 
 courses were laid lengthwise an«l secuied to the beams with 
 scrcw.s. 
 
 / Cornices. To the ends of the ujipcr ami lower floor beams 
 were secured strong cornices. 
 
 3d. Truss rods. The truss rods were at first made of 1 in. 
 diameter round iron with a nut on each end. Tliev extended 
 each way from the bottom ol each jiost diagonally to the top 
 of the fourth post distant, passing lhron<jh the chords and 
 secured by the nuts being screwed down to the face of a cast- 
 iron angle washer. 
 
 4th. Susj)enders. The lower cable suspenders were attached 
 by means of stirrup to the ends of the lower floor beams. 
 
 Those of the upper cables were attached to the upper floor 
 beam on each side, about midway between the track and the 
 trusses. 
 
 5th. Stays. 
 
 a. From the top of each tower, where one end of each was 
 
■ 
 
 (i 
 
 10 
 
 secured to the saddle, sixteen wire rope stays extended to 
 various points of the upper and lower floors. They were of 
 wire rope 4^ in. circumference. The longest reached to a 
 point 250 feet from the end of the truss. 
 
 b. Attached to each side of the lower floor, at intervals of 25 
 feet for the whole length of the bridge, excepting 75 feet at 
 each end, were wind stays whose other ends were fastened to 
 the ro'^ks on each side of the river. There are fifty -six of them 
 in all. 
 
 CHANGES IN TEE OLD WORK. 
 
 After the bridge had been in use for some time, it was found 
 that the lower floor did not assist the chords until the latter 
 had been overstrained, after which the constant working of the 
 joints at the intersection of posts with bf;ams and chords, had 
 produced rapid decay of these parts. As they became weak- 
 ened and the stress came upon the floor planking, the screws 
 holding them to the beams were loosened, allowing moisture to 
 get in aid causing the upper edges of the beams to rot. 
 
 To remedy this difficulty, auxiliary chords, composed of 
 heavy timber, were bolted to the undersides of the lower floor 
 beams, parallel with, and just inside of the truss rod chords. 
 
 To obviate the necessity of removing the old beams, their 
 projecting ends were cut off, the posts were cut off level with 
 the tops of the beams, the portion between the beams removed, 
 and pieces of 6 in. x 13 in. pine timber inserted from each 
 side of the bridge, their inner ends meeting at the center line 
 and their outer ends projecting to form an attachment for the 
 lower suspenders and to which the lower cornices were attached. 
 The lower ends of the posts then rested upon the new bejims 
 and oak shoulder-blocks were spiked to the sides of beam and 
 post. The bolts holding the auxiliary chords passed up 
 through these new beams and remaining portions of the old 
 beams (between which the new ones were placed) were secured 
 by bolts passing through all three. 
 
 Such was substantially the arrangement of the parts of the 
 bridge up to the year 1877. But the upper floor had been 
 renewed in 1873, though no alteration was made in the 
 arrangement. 
 
11 
 
 When in the condition above described, the whole suspended 
 weight, between the towers, including cables, stays andsus- 
 penders, was estimated at 1,180 tons. 
 
 INSPECTION. 
 
 In the latter part of February, 1877, Mr. Thomas C. Clark, of 
 the Phoenixville Bridge Co., with a view to examining the con- 
 dition of the portions of the cable strands embedded in the 
 masonry, caused a small excavation to be made near one of the 
 shoes. On reaching the first strand, two or three of the wires 
 were found to be corroded quite through and others were 
 partially corroded. 
 
 Shortly afterward Col. W. H. Paine, of the East River 
 bridge, visited the bridge and gave orders for the removal of 
 all the masonry covering the strands of each cable. He also 
 made tests of the elongation of the strand portion of one of 
 the cables, by means of a vernier scale. He found in this 
 way that the elongation under a given moving load, on the 
 bridge, was no greater than the modulus of the wires would 
 allow, supposing the total section to be the same as when the 
 cables were new. He also cut out some pieces of wire and 
 tested them for tensile strength, ductility, &c. 
 
 Their ultimate strength was fully equal to that of the new 
 wire per unit of section, and their reduction of ruptured 
 section was satisfactory, but as the wires tested were etched in 
 places, of couise the stretch would be principally confined to 
 the etched portion, hence rendering any measurement of the 
 stretch a matter of extreme diificulty. 
 
 On the 16tli of March, 1877, the writer joined Col. Paine at 
 Suspension Bridge to assist in examining the condition of the 
 bridge and in repairing the defective wires. 
 
 After the strands were thoroughly cleaned and the wire 
 bands removed, they were opened, the paint removed from the 
 interstices and the inner wires examined. They were found to 
 be in as good condition as when first put in. 
 
 The outer defective wires were cut away so as to uncover 
 the second layer of wire at the bend of the shoe, when the 
 second layer, or course, was found to be sound and bright. 
 
12 
 
 Thus it was found that the only wires affected were the outer 
 wires of the outside strands. Near the cylindrical portion of 
 the cables, the outer wires were slightly rusted clear around the 
 cable, but as we approached the shoes, the etching appeared to 
 work toward the lower strands, till, when the shoes were 
 reached, the principal corrosion was of the outer wires under- 
 neath the bottom shoes. 
 
 The evident cause of this corrosion was the elongation and 
 contraction of the strands under the passing loads, which had 
 loosened the cement from the outside strands, allowing moisture 
 to work in and finally reach the lowest point. The portion of 
 cement among the strands would go and come in a body with 
 the;n. 
 
 Occasionally a limestone spawl had been carelessly left in 
 contact with a strand, when they were being covered, in which 
 case the contact was indicated by a black spot. On cleaning 
 this away the wires immediately under it were found to 
 be corroded partly through the outer layer. 
 
 While the examination was going on, the defective wires 
 were cut out and new ones spliced in under strain. The 
 greatest number of wires that required repairing at one end of 
 any one cable was sixty-five. 
 
 On the 29tli of March, 1877, Messrs. Milnor W. Eoberts, T. 
 E. Sickles and Col. W. H. Puine visited the bridge, at the re- 
 quest of the Board of Directors, and made an examination of 
 the cables, both at the strands and at the saddles. These gen- 
 tlemen reported that, in their opinion, as soon as the repairs, 
 then going on, were completed, the cables would be in good 
 condition. 
 
 A copy of their report was sent to the G. W. R y Co., whose 
 only reply was to call for a Commission of engineers to make 
 an examination of the bridge and report upon its condition 
 according to the terms of their contract. 
 
 The Bridge Companies then selected Col. Paine, the G. W. 
 R'y Co. chose Mr. Thomas C. Clark, find these two gentlemen 
 gelected Mr. Charles Macdonald. The Commission arrived 
 April 17th. During that and the three following days, they 
 aiade a very thorough examination of the bridge and all 
 portions of the anchorage that were accessible. 
 
 tmmm 
 
13 
 
 ibe outer 
 artion of 
 ound the 
 pea red to 
 jes were 
 IS iinder- 
 
 tion and 
 hieh had 
 noisture 
 )rtion of 
 >dy with 
 
 y left in 
 n which 
 sleaning 
 >und to 
 
 e wires 
 1. The 
 J end of 
 
 lerts, T. 
 the le- 
 ition of 
 se gen- 
 repairs, 
 n good 
 
 , whose 
 • make 
 idition 
 
 G. W. 
 
 tlemen 
 irrived 
 3, they 
 ad all 
 
 Very particular attention was paid to the condition of the 
 anchor chains. Measurements of the sections of all the upper 
 lengths of bars in each chain were taken, and of the siaes of 
 heads and pins of the same, showing the following conditions: 
 
 1st. The outer bars of each set had a rather greater section 
 than the intermediate ones, and the total section of the bodies 
 of the nine links of a chain was 86.626 square inches, the 
 average width of each bar being 7 inches, and the thickness If 
 inches. 
 
 2d. The diameter of the pin was found to be scant 3^ inches, 
 or hardly five-tenths of the width of the body of the bar. The 
 form of the eye bar head was approximately circular, with a 
 diameter of about 12 inches. The ceuicrs of the pin holes 
 were distant from the centers of the heads, and toward the 
 body, so far that in some instances the minimum section on 
 each side of the eye was less than half that of the body of 
 the bar. 
 
 3d. Some of the pins through the shoes were found to be 
 bent convex toward the cable 5-16 inch in their total bearing 
 lengths. 
 
 4th. Two or three of the pin holes were open, back of the 
 pin, nearly ^ inch, while at each side of the pin they were close 
 down, showing the holes to be eliptical. This was probably 
 caused by the hole having been bored a little too large, which 
 allowed them to elongate, or the bars may not have been quite 
 long enough. At the close of the inspection a test load of 
 nearly 450 tons, composed of a switch engine and twenty 
 loaded box cars, was run upon the bridge and points of the 
 curves of undulation taken with a level for each position of the 
 load, at stations 100 feet apart. 
 
 The bridge regained its original camber when the load passed 
 off. The Commission also made many tests of specimen wires 
 cut from the strands. 
 
 After due consideration the Commission reported substan- 
 tially as follows: 
 
 That the repairs of wires, affected by rust, having been com- 
 pleted, the action of the wire portion of the cables indicated 
 that they were in good condition. 
 
 But regarding the anchor chains, it was believed that the 
 
1^ 
 
 form of the heads and size of the pins would not enable these 
 parts to transmit a greater ultimate strain upon the bodies of 
 the links than 40,000 lbs. per square inch of their total section. 
 Consequently while each of the cables possessed an ultimate 
 strength of 3,640 x 1648=5,998,720 lbs., or say 6,000,000 lbs., 
 the chain would not probably sustain a greater strain than 
 86.625 X 40,000=3,465,000 lbs. or 6,000,000—3,465,000= 
 2,535,000 lbs. less than the cable. The section of new chain 
 necessary to supply the deficiency, estimated at 50,000 lbs. per 
 square inch would consequently be 50.7 square inches. 
 
 The report was accompanied with plans for this reinforce- 
 ment and required that it should be made. 
 
 The report also suggested the renewal of the suspended 
 superstructure with iron, and submitted a general plan lor that 
 purpose. 
 
 REINFORCEMENT OF THE ANCHORAGE. 
 
 The plans accompanying the Commission's report were pre- 
 pared from data obtained from Mr. Roebling's published report 
 on Niagara Suspension Bridge. It proposed the sinking of 
 one new pit into the rock at the back end of each anchor wall, 
 with a chamber at the bottom for the reception of three small 
 anchor plates. To the middle plate were to be attached two 
 chains which were to pass vertically upward to the surface of 
 the rock and thence to follow a curve till they became tangent 
 to the line of the upper cable, and thence to pass one on each 
 side of the old chain and attach to the strands of the cables. This 
 connection was to be formed as follows: (see plate I.) Into 
 each shoe was to be fitted a cast iron block with one end con- 
 cave to fit against the pin. There were seven of these blocks 
 at each end of a cable. The outer ends of all these blocks to 
 be dressed off" in the same plane at right angles to the center 
 line of the cable at this point. Two cross bars, with a side of 
 each planed, were to rest against the cast iron blocks and have 
 their ends turned for the upper ends of the new chains to take 
 hold of 
 
 Each of the other two chains were to be made fast to one of 
 the outer plates, and passing upward, in a curve, till tangent to 
 
 ,r 
 
15 
 
 the line of the lower cable, and thence passing up, one each side 
 of the old wall, to attach to the lower cable in the same man- 
 ner as in the case of the upper cable chain. 
 
 The strains were to be applied by leaving one of the joints 
 of each chain, on the curve, about two inches nearer the center 
 of curvature, than any of the others and having the chain in 
 this position just the right length to allow of connecting to the 
 cable, then expand the chains by heat and as they lengthened 
 raise the low joint and block it with iron plates, after which the 
 cooling of the chain would cause it to contract and thus relieve 
 the old chain of a portion of its load. 
 
 MEASURExMENT OF STR^ 'NS. 
 
 Each bar of the upper length of each chain was to be subjected 
 in a testing machine to a stress equal to the permanent load to 
 come upon it and the corresponding elongation noted. The 
 same elongation to be given to them when in position. 
 
 EXECUTION OF THE WORK. 
 
 The plan above described was to be subject to such altera- 
 tions as circumstances should require. 
 
 I was selected as engineer in charge of the reinforcement and 
 arrived at Suspension Bridge September 13th, 1877, and com- 
 menced at once the removal of the earth to uncover the rock. 
 
 On getting to the surface of the rock, it was found that the 
 profiles of the anchorage given in the report of Mr. Roebling, 
 from a typographical error, was a combination of the profiles 
 of both sides of the river, that is, that the wall of the Canada 
 anchorage and the height of the rock surface on the New York 
 side were given in the same profile. Hence by placing the pits 
 at the back of the old wall in each case, the line of the lower 
 cable, on the Canada side, would enter the rock 30 feet away 
 from the pit, and that on the New York side 12 feet. This 
 would require expensive trenches in the rock, the cutting of 
 which would necessitate blasting nearer to the old anchor pits 
 than would be safe. 
 
 I accordingly made alterations in the plan which will be un- 
 derstood by referring to Fig 1, Plate I. 
 
16 
 
 In this plan the pits were located the aame as in tlie other, 
 but smaller. One anchor plate in each pit was made to answer 
 for all the four chains. There were eight links secured in the 
 plate by one pin, and the first joint (c) was secured by one long 
 pin. Beyond (c) each of the four chains was independent of 
 the others, but hcjd the same curvature and rested on the same 
 sione supports. Two of the chains connected with the upper 
 cables. The other two passed along grooves cut in each side 
 of the wall, passmg the supports of the old upper cable chains 
 and fastened to the lower cable. 
 
 As will be seen by Fig. 1, Plate I, the plan followed, required 
 a bend in the lower cable chain, to bring it on to the line of the 
 cable. This was done by dividing the change of direction 
 among three points, and securing them in position by means of 
 stirrups attached to the ends of the pins of the old chain, as 
 shown at a, b and c. 
 
 The connection with the strands was made as in the first 
 plan, excepting that instead of the last links attaching directly 
 to the cross bars, a triangular link is interposed, as shown by 
 the dotted line in Fig. 2. The arrangement of the holes in 
 this I'nk is shown in Fig. 3. The object of the link is to cause 
 a proper distribution of the strains. There b^ing three stiands 
 above and four below, three-sevenths of the strain should go to 
 the upper, and four-sevenths to the lower strands. Hence the 
 resultant of the strain on the entire chain was made to coincide 
 with the center line of the cable at this point, e and J] the 
 centers of the upper and lower cross bars respectively, are in 
 a plane at right angles to the center line of the cable, and the 
 line joining e and/ is intersected by the center line at a point, 
 distant from e, four-sevenths of the whole distance, a. d, the 
 point of attachment of the chain to tho triangular link, is on 
 the center line of the cable. 
 
 APPLICATION AND MEASUREMENT OF STRAIN. 
 
 These were made as before described in the case of the 
 upper cable, but in that of the lower it was not necessary to 
 apply heat, as the joint at A could be raised up and blocked, 
 after which the strain was effected by screwing down the nuts 
 
17 
 
 of stirrups a, b and c (Fig. 1) until the scale irdicated the 
 proper elongation of the eye-bars. 
 
 The total section of new chain attached to each cable is 50 
 inches for all that portion above the curve. From the upper 
 end of the curve downward it gradually decreases till, at the 
 anchor plate, it becomes 40 square inches. 
 
 METHOD OF DOING THE WORK. 
 
 1st. Sinking the Pits. — In plan the pits are 6 ft. x 2 ft 
 6 in. On the New York side they were sunk to a depth of 17 
 feet On the Canada side to 23 feet At the bottom the pits 
 were chambered to 6 f t x 7 ft in plan, for the reception of the 
 anchor plates. 
 
 In sinking the pits holes were first drilled along the four 
 sides of each as near together us the drills would work 
 without running the holes together, and to nearly the full 
 depth of the pit The core was then blasted out with light 
 charges of dynamite. 
 
 The roofs of the chambers were dressed true and bush ham- 
 mered. Just above the chamber notches were cut in the sides 
 and ends of the pit 
 
 2d. Anchor Plates.— These are of cast iron 5 ft 6 in. 
 square and strongly ribbed. Each plate has eight cavities 
 cored into it for the reception of the lower heads of the links, 
 enclosing them perfectly. One pin passes through the whole 
 eight links and all the partitions of the plate. The upper sur- 
 face slopes outward and downard each way from the link 
 openings. 
 
 After the plate was properly placed in the pit it was solidly 
 concreted underneath. The stone blocks above the plate were 
 cut to fit each place with thin joints, and the pieces as large as 
 could be got into the chamber and notches. All vacant places 
 were filled solidly with stone and cement, but no stone was 
 permitted to come in contact with the chains. 
 
 In removing the portions of the walls necessary to place the 
 new chains, they were taken down nearly to the positions for 
 the beds of the knuckle supports, after which the beds were 
 
18 
 
 prepared with poims and bush hammer in order to give solid 
 beds and thm joints. 
 
 The work of cutting the grooves each aide of tlie walls for 
 the reception of the lower cable chains required extreme 
 caution, especially where they passed along the aide of, and 
 and partly under the ends of, the stone supports of the old 
 upper cable chains. Tn the case of the south wall, on both 
 sides of the river, these old supports had been solidly bedded, 
 but in both north walls we found large cavities under two of 
 them. In such cases the cutting had to be suspended till 
 cement mortar could be forced under and have time to set, as 
 of course the least settlement of any of these supports would 
 destroy the adjustment of the cable if it did not endanger the 
 bridge. 
 
 After the new chains were adjusted the masonry was rebuilt 
 and both new and old chains covered and grouted solidly. But 
 the wire strands were covered with brick houses. Each house 
 is provided with a hatchway in the roof to give access to the 
 strands for purposes of inspection and painting. 
 
 I 
 
 DURABILITY. 
 
 Possibly the question may be asked : Will these new chains 
 continue to do their share of the work ? 
 
 By referring to the drawing (Fig.l, Plate I) it will be seen 
 that the only manner in which the chains can become slack, is 
 from settlement of the knuckle supports and shrinkage of the 
 joints around the anchor plates. But these joints are all thin 
 and the mortar had become thoroughly set before the strain 
 was applied. 
 
 At the time the strain was applied, the sun was shining 
 directly upon the chains, making them warmer than they can 
 possibly become again, covered with masonry as they now are, 
 and hence the strain would naturally increase by covering 
 them. Again one of the upper cable chains was left uncovered 
 from D to F and without any support at E for two weeks. That 
 joint settled only half an inch out of a right line and remained 
 so during the two weeks, notwithstanding that the writer 
 
 
19 
 
 give solid 
 
 walls for 
 I extreme 
 ie of, and 
 »f the old 
 1. on both 
 y bedded, 
 er two of 
 nded till 
 to set, as 
 •ts would 
 inger the 
 
 IS rebuilt 
 flly. But 
 ch house 
 33 to the 
 
 w chains 
 
 be seen 
 slack, is 
 e of the 
 all thin 
 e strain 
 
 shining 
 hey can 
 low are, 
 overing 
 covered 
 s. That 
 mained 
 writer 
 
 jumped upon it and shook it in order to overcome the friction 
 of the joint 
 
 In concluding this report on the anchorage, it is proper to 
 speak of the developments made by uncovering the old chains. 
 
 From the condition of the cement around the curved portion 
 of the chain it appears that the miisonry covering them in the 
 wall was put on before the weight of the bridge came upon 
 them As this weight was added the curved portion of the 
 chain had settled forward toward the river and downward. This 
 settlement was greatest near the top of the wall. From this 
 point downward it decreased till about the third joint down 
 where it was very slight. That such settlement had occurred 
 was shown by there being cavities behind the chains and ends 
 of the pins, both of which were rusty, while in front of the 
 chains and pin ends the cement was close, and when chipped off 
 left the iron bright as when new. 
 
 The rust was all scraped off and the old chains painted, 
 after which they were re-grouted and c<^vered by the masonry, 
 so that no water can reach them again. 
 
 I neglected to mention, in its proper place, that the cross 
 bars resting against the cast blocks in the shoes are of crucible 
 steel, of a pretty high grade, but thoroughly annealed. There 
 was not sufficient room between the strands to properly place 
 iron bars of sufficient size to afford the proper strength and 
 stiffness. 
 
 RENEWAL OF THE SUSPENDED SUPERSTRUCTURE. 
 
 While the work of reinforcing the anchorage was in progress, 
 I made a careful study of the old superstructure, both to ascer- 
 tain its condition and also the requirements of a new iron truss 
 that would be adapted to take its place. 
 
 The" Condition of the Old Trusses. — About they ear 
 1873, the upper floor had been renewed. The lower had, in 
 addition to the auxiliary chords and slip beams (already men- 
 tioned), been repaired from time to time, new pieces being 
 inserted, here ahd there, as the old ones had become decayed 
 or broken. 
 
 r.7xsr.'?a<°.'.'a" 
 
20 
 
 As the strain from the trusses had to be transmitted to the 
 auxiliary chords by the transverse stiffness of the floor beams, 
 the frequent wrenching, to which these latter had been sub- 
 jected, had opened their tibre, and, from the action of water 
 and frost, had caused them to decay with excessive rapidity. 
 
 The old truss rod chord had become nearly worthless before 
 the auxiliary one could act, and vfhen the latter did begin to 
 act it was compelled to do the whole work, consequently it had 
 already parted in several places. 
 
 In short, while it appeared that by constant repairing the 
 structure had been kept in safe condition, it was evident that 
 the expense of keeping it so must greatly increase, and that, at 
 no very distant date, it would become necessary to renew the 
 entire truss system either in wood or iron. 
 
 Whatever course should be decided upon, it was necessary 
 that the work should be done without stopping the traffic. To 
 effect an entire renewal, with either wood or iron, without stop- 
 ping the traffic, would require an entire change from the old 
 plan. It was desirable to decrease the weight as much as pos- 
 sible. To do so with timber would require the best of sounds 
 clear material, which is every year becoming more scarce and 
 expensive. Even then it is doubtful whether a wooden 
 structure, possessing the requisite strength, could be made of 
 less weight than the old one on account of the loss of strength 
 caused by cutting away material to form splices suitable to 
 resist tensive stresses. Moreover the best constructed wooden 
 structure would require renewal again in a few years. 
 
 The manufacture of iron had reached a degree of perfection 
 that rendered it a cheap material, and it was open to none of 
 the objections pertaining to wood. 
 
 CONDITIONS TO WtllCH THE NEW SUPERSTRUC- 
 TURE MUST BE ADAPTED WHEN MADE OF 
 IRON. 
 
 1st. The object of the trusses is to prevent too great undula- 
 tions of floors and cables from the action of partial live loads 
 and from wind, but yet it must not possess too great rigidity. 
 
 liMwuwim-iMijuaiiMiajiaH i 
 
 mm 
 
21 
 
 2d. The cables being anchored at both ends, there is, between 
 extremes of temperature, a difference in versed sine of about 
 two feet. The trusses must be arranged to accomodate them- 
 selves to that change without taxing the cables too much in 
 deflecting them. 
 
 8d. If stays from the towers to different points of the floors 
 contmue to be used, the trusses should be continuous, and the 
 middle point of their length should have as little movement 
 endwise as possible in order to make the stays effective at all 
 temperatures. The only use of the stays is to assist the trusses. 
 
 The first of the above conditions is met by giving to the 
 trusses a depth such, that with the maximum deflection, allow- 
 able, under a maximum bending load, the members will not be 
 subjected to more than the proper stress per square inch. 
 Then give to each member such a section as v/ill enable it to 
 resist greater deflection. 
 
 The second requirement is satisfied in a manner similar to 
 the first, that is by giving the trusses a depth that will enable 
 the cables to bend them half the amonnt of their change of 
 versed sine without too great stress upon cables or truss mem- 
 bers Probably, in most cases, if this condition is satisfied, the 
 depth will be found to be right for the first condition. 
 
 The third condition may be satisfied by making the trusses 
 continuous from end to end, while at each abutment is arranged 
 some means for automatically limiting the movement endwise 
 to a given p.mount. 
 
 In the present case the depth of truss was approximately 
 fixed by that of the old structure. But the depth was reduced 
 as much as possible by placing the lower chord on top of the 
 lower floor beams instead of below them as in the old structure. 
 
 While the work of reinforcing the anchorage was going on, 
 I prepared plans for renewing the suspended superstructure in 
 iron. They received Col. Roebling's approval, and it was my 
 wish to let the contract for furnishing the materials so that it 
 could be delivered during the fall and winter of 1878-9 and be 
 ready for erection in the following spring. But the Board of 
 Directors was not ready to commence upon it at that time and 
 nothing more was then done about it. 
 
• I 
 
 if 
 i ■ 
 
 ! 
 j 
 
 pi! 
 
 22 
 
 During the following winter the question of using steel for 
 the suspended superstructure of the East River Bridge came 
 up and I turned my attention to obtaining what information I 
 could regarding that material, to determine its adaptability to 
 this work. I here take occasion to acknowledge my indebted- 
 ness to Col, W. n. Paine for many valuable notes and sug- 
 gestions upon the use of steel for structural purposes. 
 
 There appeared to be no doubt as to its being the best mate- 
 rial for many of the members of the bridge, provided that suit- 
 able shapes could be obtained, as its great strength would admit 
 of decreasing the dead load of the bridge materially. But the 
 use of steel had not yet reached a point at which all the 
 required shapes could be obtained economically. 
 
 Consequently I prepared specifications which would be 
 adapted to the use of either iron or steel, or a combination of 
 both materials. 
 
 During the discussion of the steel question I submitted my 
 general plans to the members of the Commission for their ap- 
 proval which was given in a letter to the Board of Directors. 
 
 On the 7th of May, 1879, a committee on construction, ap- 
 pointed by the Board of Directors, met at the bridge office, and 
 after discussing the plans and specifications, authorized me to 
 invite lenders, from manufacturing firms, for furnishing the 
 materials and erecting the same. 
 
 May 28th the bids were opened in the presence of the Board 
 and the contract awarded to the Pittsburgh Bridge Co., that 
 company being the lowest bidder. 
 
 By the specifications and the terms of the contract the 
 materials were all to be delivered at Suspension Bridge, ready 
 for erection, by August 1st, 1879, and the erection completed 
 by November 1st, 1879. 
 
 The specification gave a preference to Bessemer steel, made 
 under the " Hay Process," for the reason that it had been the 
 material used, exclusively, in a bridge built at Glasgow, Mo., 
 by Wm. Suoy Smith, and which had given very good results. 
 But the Pittsburgh Bridge Co.'s tender was for Bessomer steel 
 and for iron. Bessemer steel met the requirements of the 
 specifications very well as will be seen by the following ex- 
 tract from that document : 
 
 mttmummmmmm 
 
4^ 
 
 ? Steel for 
 itlge came 
 )rnintion I 
 tability to 
 indebted- 
 and sug- 
 
 ■ 
 
 Jest mate- 
 that suit- 
 uld admit 
 But the 
 ii all the 
 
 ^'ould be 
 ination of 
 
 itted my 
 their ap- 
 rectors. 
 2tion, ap- 
 ffice, and 
 d me to 
 bing the 
 
 le Board 
 Co., that 
 
 •act the 
 
 , ready 
 
 mpleted 
 
 1, made 
 een the 
 w, Mo., 
 results. 
 ST steel 
 of the 
 iDg ex- 
 
 
 23 
 
 QUALITY OF MATERIALS REQUIRED. 
 
 All wrought iron must be of a tough, fil)rou8 character, 
 double relined. Burs, of moderate dimensions, must be capable 
 of being bent cold, through 90**, with un inner diameter of cur- 
 vature, not ttxceeding the thickness of the bar, without crack- 
 ing. When nicked and beut sharply over the corner of an 
 anvil, the fracture produced must show a good clean fibre. It 
 must have an ultinmte tensile strength of at least 60,000 lbs. 
 per scjmvre inch of sectional area, and an elastic limit of at least 
 25,000 lbs. per square inch, when tested in specimens turned to 
 cylindrical form, the turned portion of which has a length of 
 at least ten diameters. It must elongate at least 15 per cent, 
 before rupture. The area of action at point of fracture should 
 red ice as nearly 26 per cent, of original area as possible, in 
 order to be uniform in hardness. A uniform modulous is 
 desirable. 
 
 The steel must be of a quality such, that in dimensions to 
 be used, it shall have an ultimate tensile strength of not less 
 than 70,000 lbs. per square inch ot section, and an elastic limit 
 of not less th>tn 40,000 lbs. per square inch in tension, or in 
 compression, when the specimens have a moderate length com- 
 pared to their diameters. 
 
 Specimens bent cold through 90®, with an inner diameter of 
 curvature not greater than one and a half diameters of the bar, 
 must not show any cracks or flaws. When the specimens are 
 tested in tension, with a strain of 40,000 lbs. per square inch of 
 section, their power for resisting shocks may be tested by 
 striking them smartly with a hammer while under that strain, 
 when they must not crack. 
 
 The specimens must elongate at least ten per cent, of their 
 original length before rupture, and their ruptured sections must 
 be reduced at least twenty per cent, of their original sections. 
 Uniformity in modulus and ductility is to be secured as far as 
 possible. 
 
 On reaching Pittsburgh I made some tests of specimens of 
 Bessemer steel. They gave very good results ; but the only 
 shapes to be obtained in steel were plates, angles and small 
 channels. 
 
H 
 
 24 
 
 I then decided to use steel for the posts, chords, track string- 
 era a"d lateral rods. All other parts to be of iron. 
 
 As the chords must be constructed of plat3s and angles 
 riveted together, the net section would necessarily be reduced 
 to a greater extent than if channels were used, on account of 
 the greater number of rivet holes required, hence there could 
 be no great advantage in point of weight by using steel, unless 
 of a higher quality than that called for by the specifications. 
 
 I then found at what increase of price per lb. the Hay steel, 
 with an ultimate strength of 80,000 lbs. and elastic limit of 
 48,000 lbs., could be obtained and decided to use that material 
 for the chords. 
 
 PROGRESS OF THE MANUFACTURE. 
 
 ,::i 
 
 N 
 
 Immediately after the contract had been awarded for furnish- 
 ing this material, there started up a demand for iron and steel 
 that was unprecedented. The mills, throughout the country, 
 had more orders than they could fill. Consequently when the 
 time arrived, at which the materials should have been delivered 
 at the bridge, the work at the shops had scarcely commenced, and 
 very little of the raw materials had come from the mills. It 
 soon became evident that nothing could be done about erection 
 during the fall of 1879, However the work generally went 
 forward in the shop when there was sufficient material on hand 
 to work to advantage. 
 
 Every precaution was taken to do the work in a satisfactory 
 manner. 
 
 Tests were made upon specimens cut from every blow, and 
 then from specimens cut from the actual shapes from each blow 
 for all the steel used, and with no other preparation. (See 
 appendix.) 
 
 In building the chords, the angles were first laid out for 
 punching. Tue. holes were punched about ^ iixoh too small. 
 They were then clamped to the plates, in proper position, and 
 a drill of proper size passed through both angle and plate, the 
 angle being thus reamed and at same time serving as a templet 
 for drilling the plate. This prevented any possibility of one 
 
 mm 
 
25 
 
 part bringing any stress upon another, as the holes perfectly 
 agreed. 
 
 The rivets were all driven by pressure. They were of steel 
 in all the steel parts, but of a lower grade than the pieces in 
 which they were driven. 
 
 For further information upon riveting and tests see appendix. 
 
 ERECTION. 
 
 In March, of the present year, I returned to Suspension 
 Bridge to make arrangements for erection. 
 
 An examination was made of the condition of the lower 
 floor beams. From this examination it was evident that the 
 beams were not in a suitable condition to sustain the extra 
 weight that would come upon them while the new work was 
 being erected, from the fact, that in addition to the natural de- 
 terioration they would require deep notches cut into them, thus 
 weakening them still further. 
 
 Consequently I decided to put the new iron beams in, nearly 
 throughout, before commencing the work of erection proper. 
 
 A commencement was made upon this April 13th, 1880, but 
 it did not go forward rapidly until the 21st of that month. 
 The work began at the middle and proceeded toward each end. 
 As it went forward the floor was taken up for a width of about 
 3 feet on each side and a temporary wooden railing was put up 
 on each side of the carriage way, leaving, for that purpose, a 
 space of 9 feet in width. 
 
 As soon as an old beam was taken out an iron beam was put 
 in its place, the suspenders attached to its ends, and the old 
 chords and posts firmly secured to it. 
 
 This part ot the work was done by May 13th. On the 31st 
 of May, most of the material having arrived, we started on the 
 work of erection. At that time we had stripped the old bridge 
 of all superfluous material, as cornices, the rails forming the 
 broad gauge track, &;c., in all decreasing the weight about 
 80 tons. 
 
 METHOD PURSUED IN ERECTING THE NEW WORK. 
 
 Plate IV is given to show the respective positions of the old 
 and new work. 
 

 I,. 
 
 1 1 
 
 ill 
 
 i ■ ' 
 
 26 
 
 The new superstructure is made as wide as would admit of 
 its being placed between the rods of the two old trusses. The 
 bottoms of the iron lower floor beams were placed at the same 
 level as those of the old beams. The tops of the new upper 
 beams were about an in inch below the tops of the old ones at 
 the ends. 
 
 The steel upper chords are 3 inches narrower than the lower 
 ones, in order that, by cutting away 3 inches in width from the 
 inner edges of the wooden chords and notching the tops of the 
 wooden floor beams, the steel chord could be placed in poHtion. 
 
 The outer channel of the lower chord came in the position 
 of the foot of the inner hall ot the old post. 
 
 The erection proceeded as follows : 
 
 1st. Blocks of pine 4 in. x 6 in. and of various lengths were 
 prepared and tacked, one to the foot of each post, (the lower 
 ends of the old posts consisted of blocks of red cedar.) 
 
 2d. Beginning at the middle of the bridge five lengths of 
 steel lower chord were placed in position, on each side of the 
 lower floor. The red cedar blocks being removed to make room 
 for them, and as soon as a length of chord was placed, the new 
 4 in. X 6 in. blocks were inserted between the foot of the inner 
 half of the post and top of the steel chord. 
 
 As the chords were placed their splices were riveted up. 
 The rivets for this purpose are of steel from f in. to ^ in. diam- 
 eter. They were driven while hot, with 8 to 10 lbs. sledges, 
 which had the effect of upsetting the rivet and making it fill 
 the hole nearly as well as by pressure. 
 
 3d. Beginning at the middle again, one part of each upper 
 floor beam was cut off and the middle portion taken out to 
 make room for the iron beams. These were inserted by passing 
 them through between the old track stringer, turning them up 
 on edge and placing them. As soon as the iron beam was in 
 place a pair of steel posts were set up, the lower chord pins 
 passed through the chord, the foot of the posts, and the eyes of 
 the truss rods stubs. The heads of the two posts were then 
 placed under the ends ol the upper beam and temporarily 
 bolted to its lower flanges. The beam was then wedged fiimly 
 between the parts of the old track stringer with wooden 
 wedges. 
 
 1 
 
27 
 
 d admit of 
 isses. The 
 it the same 
 new upper 
 )ld ones at 
 
 the lower 
 li from the 
 ops of the 
 n pof'ition. 
 5 position 
 
 gths were 
 the lower 
 
 3ngths of 
 ie of the 
 ike room 
 the new 
 ;he inner 
 
 Jted up. 
 n. diam- 
 sledges, 
 g it fill 
 
 3 upper 
 1 out to 
 passing 
 lem up 
 was in 
 rd pins 
 eyes of 
 e then 
 orarijj 
 fiimij 
 'ooden 
 
 4th. As soon as twelve of the new bents were placed, three 
 lengths of upper steel chord were placed in position, on each 
 side and bolted to the top flanges o( the iron beams. Their 
 splices were then riveted. 
 
 5th. The new truss rods were then inserted. They are in 
 pairs, made of 1^ in. square iron. Each rod is in two pieces. 
 Each piece has an eye at one end and a thread cut on the other. 
 The piece in the lower chord is short and has a left hand thread 
 cut on it. The two parts are connected and adjusted by a 
 sleeve nut. Each rod extends from the foot of one post to the 
 top of the third post from it, and they incline each way. 
 
 In erecting, but one rod was put in each place at first in 
 order to save weight. 
 
 In the above order the work proceeded each way from the mid- 
 dle of the bridge toward each end, 
 
 6th. When 150 feet of the new work was in place, the new 
 chords were securely clamped to the old by means of oak and 
 pine timber. 
 
 In the case of the upper chord the timber was notched to the 
 pin ends, rivet heads, &c., at the side of the steel chord and 
 resting on the top of the wooden chord. The upper side of the 
 wooden chord and the under side of the timber were notched 
 at intervals for the reception of oak keys. The timber was 
 then clamped strongly to both old and new chords, and the 
 keys inserted in the notches. 
 
 The old and new lower chords were clamped by fitting 
 pieces of 4 in. x 16 in. pine to the lattice straps and rivet heads 
 on the under side of the steel chord. It was keyed to the top 
 of the auxiliary chord. The whole was then bolted together. 
 
 7th. As soon as the clamps were secured, the work of renew- 
 ing the old material was commenced at the middle of the 
 bridge, and followed that of erecting the new, generally, with, 
 an interval of about 75 feet each way. 
 
 The portion of the new work thus put in place weighed 
 about 1,100 lbs. per running foot of bridge. Hence there were 
 seldom over 90 tons of new material overlapping the old, but 
 at the start, being in the middle, this was equivalent to about 
 150 tons distributed, or deducting the 80 tons, saved by strip- 
 
28 
 
 'I 
 
 m 
 
 f 
 
 M'' 
 
 ping the bridge, we have lef « 70 tons as the probable extra dead 
 load upon it, but as the trains had at the outset been limited 
 to 190 tons, it is not probable that the total weight of live and 
 dead load ever exceeded that of ordinary usage. 
 
 While the above rhanges were being made, the work of re- 
 placing the lower floor was going forward each way from the 
 middle. As the planking of the old floor was laid longitudi- 
 nally, while that of new is laid transversely, there was neces- 
 sarily a gap between them, which was bridged over by a raised 
 platform, under v/hich the change was made, and which was 
 moved along as the work progressed. 
 
 When the work of making the above changes had reached 
 the towers that of completing the riveting was carried through. 
 It was all completed, except changing the track, by the 25th 
 of August. Up to that time the traffic had not been inter- 
 rupted. 
 
 10th. After the work of replacing the trusses and floors was 
 completed, that of renewing the track began at the middle and 
 proceeded each way at the rate of 30 feet per day, or of 60 feet 
 total. This could have been done without interrupting traffic, 
 but as the Great Western Hallway Company was to do the 
 work of removing the old material of the track and put on the 
 new timber, they preferred to take an hour each day, when 
 there was no passenger train and scarcely any freight to cross, 
 and make the change of 60 feet at one time. This gave the 
 remainder of the day to preparing more old material for re- 
 moval, and for completing the riveting of the new. 
 
 The whole change was completed September I7ih. Since 
 the completion of the changes above described the time has 
 been occupied in making side walks, hand rails, putting on the 
 new roof, making new fastenings for the overfloor stays at the 
 saddle, on the towers, and in painting. 
 
 The brick houses have also been built to cover the strands 
 at the ends of the cables. 
 
 ADJUSTMENTS. 
 
 The work of adjusting the new structure has been carried 
 forward as rapidly as circumstances would permit. 
 
 ■■an 
 
29 
 
 The operation was as follows : 
 
 Ist To Adjust the Camber. — It was desirable to make it as 
 nearly an arc of a circle as possible. Consequently the truss 
 rods were slackened in order to allow the cables to assume their 
 natural curves. Levels at each 100 feet, in the length of the 
 bridge, jvere taken to ascertain its profile while hanging 
 naturally, then the true arc, with a proper versedsine for mean 
 temperature was calculated, after which the difference between 
 the true arc and the actual camber was found for each of the 
 points. The suspenders were then screwed up or unscrewed 
 according as the structure required raising or lowering till the 
 proper camber was given. As this raising or lowering pro- 
 gressed it was necessary to keep the rods slackened. 
 
 2d. Stress on Suspenders. — This adjustment was made 
 by means of a hydraulic weighing machine. A portable 
 wooden frame was made to rest against the under sides of the 
 beams. One end of the machine was secured to a screw passing 
 through a cross beam of this frame while the other end of the 
 machine w..s attached to the screw end of the suspender. The 
 nuts on the screw ends were then slackened off". Then by 
 working the screw at the other end of the machine the right 
 strers was given to the suspender as indicated by the index. 
 
 The first time of going over them the machine was attached 
 to one of the suspenders and the proper stress given to it. 
 Then, leaving the machine in that position, the five or six sus- 
 suspenders on either side of it were adjusted by springing them 
 laterally with the hand, using the suspender with the machine 
 on it, as the standard. After going over them in this way from 
 one end to the other the machine was tried upon an occasional 
 suspender to verify the stress. This required to be gone over 
 three times to bring them to proper adjustment. 
 
 3d. Truss Rod Adjustment. — The truss rods required to be 
 adjusted at about mean temperature, at which time there should 
 be no stress upon any of them when the bridge is unloaded. 
 
 Consequently at about mean temperature each of a pair of 
 rods was first screwed up to an equal bearing and then the 
 sleeve nuts were turned backward just half of a revolution. 
 This gives the trusses a little more flexibility, decreasing the 
 
1$ 
 
 SO 
 temperature stresses and Af fi 
 »°«than the proper amoU or^'r":' '^"^^ -'allow of 
 
 were adjusted by ru„„i„g , ^f^/"";; f.^ndred feet ou-, they 
 «.- -e.i„, e.h or thf .a/s'^tolfpr^X.^ "" ""^ 
 mSENT ARBANGEMENT OE THE BRIDGE 
 
 -HorrHr:^^::^^-,:-;^; e.eept.-o„ on.e 
 
 ■Plates II and Tfl »h / described. 
 
 But the action or so „e If 1?™"^^"'^"' "^ ">« t^ss system 
 
 «on bndge of this sort, i„ order! , T ?'"''' '" "^ ™^Pen- 
 
 (or those from the tops of th^T ''^ ""^ """'^o^' stays 
 
 floors) effective, a con'ltt ' rruf .'^''^«'»' ?"■■"« of the 
 
 point of whose length shall bTsl", '' r^"^- '^e ".iddle 
 
 The trusses in this ease are cont Ln!!' V "'""'■'^ "'" Po^^iW"- 
 
 order to keep the middle ftom mo "" *"<' '" «"d. I„ 
 
 automatic device shown at "he ZdJ .'"?''' "'"'«' '"^ 'he 
 
 n and m, „,, designed. Its ^L ^ '""''' '=''"''• ^'^'es 
 
 In the prolongation of the „e of t." f"^"^^ '' ''""""s : 
 ment casting A (Plate Iliriirmlv ' T' "^""^ '» »" "but- 
 'he arch. This casting received the T'^'" '^' ™^«°'»'y of 
 There is one of these fast n's , tch'e ;';"' "' ">« «''"'». 
 A bent lever B ha^ .-.s fulcrum fit ^ '" '"*«■• "hord. 
 P) of the short arm ... the leve^ ;! t""'''''' '^ ^- A' 'he end 
 diameter round rod R Th i, V "^^"^ °"^ ^"d of a f in 
 chord to the oppositf sid?:f Ihe ^'^"'^^"-^o-gh the lowe; 
 secured to the abutment castl h ' 7^'"' "' °'her end is 
 
 of .'he long arm of B is susp /ded^" ""' "^^ ^' "'<' cd (F) 
 IS mterposed between the endc^'^l^t "?" "^"^^ «• «hich 
 men. casting. The action of the d, •' "^°''" ""'^ "^ *he abut- 
 
 The change i„ length of he X ^'t "' ^"""^^ ^ 
 'emperature, is about 8^ inche If ^K ' ^'.''"'" ^^^^'"es of 
 
81 
 
 extremes. The rod R, whicli lies loosely in the chord but 
 otherwise is independent of it, is a little longer than the chord 
 and will change in length, between extremes, 8^ inches, or 
 double the movement of either end of the chord. Hence the 
 other end of the rod being fast, the end D will move 8^ inches 
 carrying the end of the lever with it at the same time that the 
 end of the chord moves 4J inches. Arm E F of the lever is 
 three times the length of D E, hence F will move 25^ inches, 
 or six times as far as the end of the chord moves. Consequently 
 the wedge C is made with an inclination 1 to 6 of its length. 
 There is one ot these wedges at each end of each lower chord. 
 When the chord contracts the rod contracts in the same pro- 
 portion and at the same time, thus bringing a thicker part of 
 the wedge between the chord and abutment. 
 
 There is half an inch of space at each end for the chord to 
 go and come in before bearing upon the wedge, an amount 
 which is very nearly constant for all temperatures. 
 
 The long rods lymg inside of the chcrd they both keep at 
 nearly the same temperature with each other. 
 
 The wedge has two g- rfaces of friction, and hence its incli- 
 nation of 1 to 6 is far within the angle of friction of cast iron. 
 Hence no matter what the pressure of the chord, it brings no 
 stress upon rod R except what is required to sustain the weight 
 of the wedge. 
 
 Should the wedge ever get caught by the chord remaining 
 against it, there being scarcely two hours, day or night, in 
 which a train does not pass, as soon as a train rests on the other 
 end of the bridge the wedge will be released. 
 
 Fig. 1, Plate VI is to show the action above described. 
 
 2d. Action of the Overfloob Stays —Fig. 1, Plate VI is 
 to assist in explaining this. 
 
 In case of a wooden truss, the points of attachment of the 
 stays to the floor move in vertical lines as the temperature 
 changes. Their change in length is not sufficient to compensate 
 for the amount that the cables move the floor vertically. 
 Hence, if they are adjusted properly for cold weather, they 
 will become so tight in summer as to break and vice versa. In 
 the case of an iron truss with a slip joint anywhere beyond the 
 
82 
 
 ill! 
 
 ifi: 
 
 
 attachment of the stays and with the end of the truss fixed at 
 the tower, the point of attachment will move in a line c d, 
 in which case its action is worse than in that of a wooden 
 truss. 
 
 But in the case of a continuous truss with the middle kept 
 stationary, the point of attachment moves in the line a b. In 
 this way the stay is made to compensate very nearly for all 
 changes of temperature. 
 
 3d. The ends of the trusses are anchored to prevent vertical 
 and lateral movement only. 
 
 At the New York end this is effected by a hinged strut 
 
 At the Canada end, where the surface of the rock is nearly 
 at the level of the lower floor, a casting, anchored to the rock, 
 is provided with slots, in which blocks on the ends of the end 
 pins of the lower chord are free to move in the direction of the 
 length of the bridge only. 
 
 4th. Suspenders. — On account of the ends of the trusses 
 being anchored to the rock the suspenders near the ends require 
 to be left slack. It would be as well if they were left off, but 
 as they do no harm they were kept on more for appearance 
 sake than otherwise. 
 
 5th. Upper Floor. — The deep transverse beams of the 
 upper floor, together with the deep longitudinal track stringers, 
 distributing the load over several beams, and the upper cable 
 suspenders attaching to the beams, as shown in the transverse 
 view, Plate II, form a very stiff floor. This makes the office 
 of the knee braces that of steadying the bents merely. 
 
 6th. Lower Floor. — The planking of the lower floor is laid 
 crosswise ot the bridge and in one thickness, to enable water 
 to drain off more readily and the floor to dry out quickly. 
 
 The narrow foot- walks at each side serve the double purpose 
 of a clean walk for pedestrians and to prevent the snow being 
 blown from the carriage-way in winter, while, at the same time, 
 they confine the portion loaded with snow to the necessary 
 width for carriages to pass each other. 
 
 7th Lateral Bracing. — As neither the upper or lower 
 floor has any longitudinal planking to afford lateral stiffness, 
 diagonal rods of steel were introduced to supply the deficiency. 
 
 MM 
 
33 
 
 The fiftysix wire rope river-stays are retained to resist the 
 action of high winds and consequently the diiigonal rods are 
 merely for the purpose of keeping the intermediate points of 
 tiie cliords in line. 
 
 STRENGTH OF THE BRIDGE. 
 
 The anchorage, cables and towers are ])rimarily the support- 
 ing members of the bridge as before. Hence the strength of 
 these members is to be considered first, in the order observed 
 above. 
 
 1st. The Strength of the Anchorage. — This has already 
 been discussed under the head of Reinforcement of Anchorage. 
 
 2d. The Cables. — The number of wires in each of the four 
 cables is 3,640. The original, average, ultimate strength of 
 each wire was 1,648 lbs. This gave, as the strength of one 
 cable, 3,640 x 1,648=5,998.720 lbs.=2,989 tons, or for the 
 four cables=2,999 x 4=11,996 tons in the direction of their 
 length. There is no indication of any deterioiation in their 
 strength, but su])pose the strength of the four cables to be 11,- 
 000 tons in the direction of their length, or 1,511 lbs. per wire. 
 
 The present total suspended weight of bridge, between the 
 towers and including cables and stays, is 1,050 tons. Taking 
 the maximum live loud upon the bridge at one time as 350 tons 
 it makes the total live and dead ]oad==l,4'^0 tons. The maxi- 
 mum stress upon the cables is at the top of the tower. Their 
 stress at this point, in the direction of their length, is to the 
 total vertical load as 1.78 is to 1. Hence we have -4:^''=6,180 
 tons. The factor of safety is then ^fSg==4.41. 
 
 This factor of safety, in a tension member as long as one of 
 these cables, and where the load, from the time it starts upon 
 the bridge till it produces its maximum ell'ect, is rarely less 
 than one minute, is ample. 
 
 3d. The Towers. — The maximum load imposed upon the 
 top of ono tower by the cables is 700 tons, and acting in a 
 nearly vertical direction. The least section of the tower is just 
 below the top and is 64 square feet. Hence the pressure per 
 square foot, from the cable, is Y/==10.94 tons. 
 
H 
 
 Trantwine gives, as the crushing load for limestone, 250 to 
 1,000 tons per square foot Hence if yi% take 250 tons as the 
 crushing load iirl4=23, nearly, is the factor of safety. Or, if 
 a factor of safety of 10 should be accepted, the number of 
 square feet necessary at the top of the tower would-»^^— =28 
 square feet, or a square whose side is 5.29 ft There is no 
 other place at which the pressure per square foot is so great as 
 at the point considered. Hence we could remove a thickness 
 of 1 foot 4 inches on all four sides of the tower and still have 
 it safe. 
 
 The stone, of which the towers are made, is limestone that 
 was quarried near the bridge. It jis very strong when used 
 where it is not exposed to moisture and frost, as for the inside 
 of a wall But where exposed as in the faces of the towers, 
 its surface disintegrates or "flakes off," and if neglected, 
 would in time work in to a depth that would endanger the 
 tower. 
 
 Painting has been resorted to as a protection but it soon 
 dries and cracks. In any case it should not be put on unless 
 the tower is dry as after several weeks of dry warm weather. 
 
 It might be well to try a coating of asphalte mixed with 
 some material to keep it from cracking. 
 
 But, in any event, the towers should be attended to and, 
 where necessary, new stones should be set in the faces. 
 
 STRENGTH OF OTHER PARTS. 
 
 SusPBNDERS. — The suspenders have not been renewed. 
 There are 628 of them to sustain a maximum load of 1,025 
 tons, or each one sustains 1.63 tons. In no case does the load 
 exceed two tons. They are of 4^ in. circumference wire rope 
 possessing when new an ultimate strength of 30 tons, giving a 
 factor of safety of at least 15. I have examined pieces of wire 
 from several of those that we have cut and tested them for 
 bending. There is no doubt of their strength being ample. 
 
 The upper cable suspenders are attached to the upper 
 floor beams by means of U shaped stirrups made out of 1^^ in. 
 round iron. 
 
86 
 
 The lower suspenders attach directly to the projecting ends 
 of the lower floor beams and close to the lower chord. 
 
 Stay Fastenings. — The stays are of wire rope, same size as 
 the suspenders. The fastenings of the stays at the towers have 
 been renewed in a manner to render the stays safe from wear. 
 
 The lower ends of the upper floor stays have permanent iron 
 fastenings riveted to the iron beams and tied to the chords and 
 track stringers by means of iron bars. The lower floor and 
 river stays are attached to the lower chord pins. 
 
 In short it has been the intention to have all fastenings per- 
 manent, in order that adjustments once made will not be dis- 
 turbed by any renewals of wood-work that may be necessary. 
 
 STRENGTH OF THE TRUSS SYSTEM. 
 
 The trusses have been designed for a maximum load of .8 of 
 a ton per foot run, and a length of load of 534 feet, using the 
 formulae given by Rankine with a factor of safety of 5. 
 
 For calculations on truss stresses, see appendix. 
 
 RESISTANCE TO HIGH WIND. 
 
 For this purpose we have the inclination of the planes of the 
 cables and the same number of wind stays as in the case of the 
 old bridge, while there is less than six-tenths as great wind- 
 surface. In a very high wind, blowing steadily, the bridge at 
 the middle swings to the leeward 5 to 6 inches and remains 
 there while the wind continues, but the motion is not felt when 
 one is on the bridge. When in that position there is more of the 
 weight of the structure thrown upon the upper windward cable 
 and the lower leeward cable, while the other two are relieved 
 of a like amount. I estimate the resistance produced by the 
 inclination of the cable planes alone in that condition at 30 
 tons. There are, at the same time, the 28 river stays on each 
 side. Those on the leeward side are relieved while the stress 
 on those of the windward side is greatly increased. As this 
 resistance is partly downward and partly toward the wind they 
 not only increase the resistance offered by the cables but will 
 safely afford a direct united resistance of 160 tons. The 8,000 
 
86 
 
 square foot of wind surface, taking the pressure of the wind at 
 50 lbs. j)cr square foot, gives 400,000 11)8 — 200 tons pressure. 
 Wliere tlie wind blows in gusts it does not appear to ull'ect the 
 bridge perceptibly. 
 
 None of the unusually high winds of this fall, though 
 blowing directly across the bridge, allected it siifliciently to be 
 felt by a person standing on it. 
 
 The strength of the trusses, together with the overflow and 
 river stays, will eflectually prevent any vertical undulation 
 from the eil'ect of wind. 
 
 WEIGHT OFTIIK NEW STRUCTURE COMPARED TO 
 
 THAT OF THE OLD. 
 
 The weight of the wooden structure, at its con]})letion, wjis 
 estimated by Mr. John A. Roebling at 1,000 tons. But at the 
 date of the inspection, there having been a large nmount of 
 titnber added to it, it was estimated to wei^h 1,180 tons. 
 
 When the work of replacing the lower iioor btams was in 
 progress I liad one of them weighed and found that owing to 
 the amount of water that it held it wa.^ veiy nmcli lieavier than it 
 had been estimated. I also weighed other pieces of the bridge 
 and from these made a new estimate with the following 
 result : 
 Total suspended weight between the j Old bridge, 1,228 tons. 
 
 towers, - - • ] New " 1,050 " 
 
 Difference in favor of new bridge, - • 178 " 
 
 It is possible that the estimate of 1.228 tons is somewhat in 
 excess. But as the new bridge is now higher in the middle 
 than the old one for the same temperature, notwithstanding 
 that the middle suspenders have been lengthened over 3 inches 
 since its completion, that would indicate a decrease of consider- 
 ably over 100 tons. 
 
 SUGGESTIONS REGARDING THE CARE OF THE 
 
 BRIDGE. 
 
 1st. As soon as dry weather comes next spring it would be 
 well to remove the hatchway covers from anchorage houses, 
 
 
 1( 
 
87 
 
 and, if tliere hIiouM he any dampnoss, let tlicm dry tlioronghly 
 uikI tluMi juiint all t;xi)f)S(Ml inm work and win-s. TluMe is not 
 much prohability of danij)no.sM as there is ample ventilation at 
 top ami hottom, but the covers should he removed once a year 
 and left off during one warm, dry day, and any part painted 
 which appears to need it. 
 
 2d. The ca|)a on top of the towers had better be removed at 
 the same time, the parts painted and niaeliine oil put in 
 around the edges of the saddles. 
 
 An examination of the cables, from time to time, will enahle 
 an experienced num to determine when tliey i-rrpiire paintijig. 
 They should be [tainted a> olten as the paint gets so hard and 
 drv Jis to crack. 
 
 2(1. The new work should be ])aintcd whenever it requires 
 it Those portions near the intersections of chords, beams, 
 posts, &c., will rcfpiire painting oftener than other parts. 
 
 3d. Spkkd of Trains. — The durability of the bridge and 
 (in case of derailment) the safety of the trains render it ad- 
 visable that some more ellicient means should be adopted for 
 enforcing the article of agreement, limiting the s])eed of trains 
 on tlie bridge to live miles per hour. 
 
 SUPERINTENDENCE. 
 
 A competent superintendent will, of course, continue to be 
 necessary. 
 
 I trust that it will not be considered out of place ior me to 
 say here, that judging from my acquain ance with Mr. W. G. 
 Swan, it is my opinion that no better man than he can ))e found 
 to fill the position. He has at all times shown himself to be 
 conscientious in the discharge of h\?. duties, is acquainted with 
 all the details of the bridge, and will therefore be competent 
 to judge, by inspection, of what is required from time to time 
 for its preservation. 
 
 In concluding this report I wish to express my sense of 
 obligation to the Bridge Companies' Superintendent, Mr. W. G. 
 Swan, for the willing assistance he has always rendered in 
 promptly supplying me with necessary tools and materials, as 
 well as for his friendly interest in the success of my work. 
 
88 
 
 Also to the Pittsburgh Bridge Co. who, in supplying the 
 iron and steel materials, called for by my plans, never spared 
 any pains necessary to secure materials and workmanship of 
 an entirely satisfactory nature. 
 
 Nor must I neglect to award great credit to my foreman, Mr. 
 William Gardner, and to the workmen, who, by careful atten- 
 tion to orders, did so much toward enabling me to complete, 
 and without an accident, a work that otherwise would have 
 been dangerous. 
 
 In resigning my position as engineer of the work, I take 
 pleasure in acknowledging to the Presidents and gentlemen of 
 both Boards of Directors my indebtedness for their kindness 
 toward me and for the unvarying confidence they have mani- 
 fested in my professional judgment. 
 
 L. L. BUCK. 
 
 P 
 $ 
 
APPENDIX. 
 
 ive mani- 
 
 CALCULATION OF STRAINS UPON THE TRUSSES. 
 
 The formulsB given in Rankine's Applied Mechanics, for 
 " Stiflfened Suspension Bridges " have been principally the ones 
 used in the following calculations, but with some modifications 
 deduced from observation of the action of the old bridge under 
 partial loads. 
 
 The formulae in question, except in one instance, treat the 
 stiffened suspension bridge as an inflexible structure, which, of 
 course, cannot be the case in pra-tice, though, by giving to the 
 trusses greater depth in proportion to their length, we could 
 approximate to the supposed condition. But it would be an 
 unnecessary condition in any case, and as before explained, is 
 not admissable in the one here considered. 
 
 The old truss system of the bridge was very flexible, deflec- 
 ting under partial live loads, in some cases as much as 2 feet 
 in a length of 500 feet. Yet the trains passed over it easily, 
 and after twenty-five years of constant use, the cables give no 
 indications of ill eflfects from the undulations. 
 
 In designing the new trusses I considered a maximum deflec- 
 tion of 15 inches in a length of 500 feet as not at all excessive. 
 Fig. 2, Plate V, is given to illustrate the advantage that is 
 gained by giving to the trusses considerable flexibility, aside 
 from the relief it affords to the cables in bending the trusses at 
 low temperature. 
 
 By allowing a deflection of 15 inches in a length of 500 feet 
 it was estimated that the intensity of the live load, per running 
 foot of bridge, and to be used in the formulse, was reduced by 
 .2 ton. In other words, if the trusses and cables were perfectly 
 flexible it, would require two-tenths of a ton per foot run to 
 
11 
 
 • 1 ' 
 
 produce the given deflection. The total intensity was taken at 
 eight-tenths ot a ton. Deducting the two-tenths of a ton we 
 have six-tenths of a ton to use in the formuJie. 
 
 The formuke are us follows : 
 
 For bending moment of trusses when the load begins at one 
 end ot the bridge and covers a given portion of it, we have for 
 the unloaded segment — 
 
 M=^^'li:-,^>JH:=^^ (1) 
 
 16 c 
 
 The moment for the loaded segment is — 
 
 M= 
 
 _W (C + X) 8 (C — X) 
 
 16 c 
 
 Max. M=Max. M='^"' 
 
 C ft ^ f 
 
 For shearing with the load as above we have — 
 
 F=^ 
 
 _W (C8 — X8) 
 
 To 
 
 Max. F= 
 
 w c 
 
 (2) 
 (3) 
 
 (4) 
 (5) 
 
 For deflection of the loaded segment when the load covers 
 two-thirds of the spun beginning at one end — 
 
 5 
 
 27 
 
 V- 
 
 1C8 
 
 e7 
 
 (6) 
 
 The formula for a distributed load, covering the whole span 
 and required to bend the two trusses 12 inches, is taken from 
 Moseley's Mechanics and is — 
 
 W= 
 
 _Dy 4 8 EI 
 
 5 C8 
 
 For chord strain produced by W, we have— 
 
 S=y^=strain per square inch. 
 
 In the above formulas. 
 W==intensity of live load per foot run=.6 tons, 
 c = half span of truss=400 feet. 
 X= distance from middle of span to end of load, 
 f = strain per square inch, on chord in lbs. 
 c = half span of truss in inches=4,800 inches. 
 E= modulus of elasticity of steel=28,000,000. 
 y = half depth of truss in inches=105.5 inches. 
 
 (7) 
 
 (8) 
 
■■ns" 
 
 D 
 
 I 
 
 d 
 a 
 
 111 
 
 = deflection at middle of truss from temperature=12 inches. 
 
 = Moment of Inertia of two trusses=l,113,000. 
 
 = length of truss in feet=800. 
 
 = depth of truss in feet=17.58. 
 
 = area of section of two chords=60 square inches. 
 
 LOAD STEAINS ON TRUSSES. 
 From (3) 
 
 Max. M=^xi^U7,lll tons. 
 Hence strain per square inch on chord section= 
 
 7111 
 
 (a) 
 
 S=i7liR^o=8 09 tons=16,180 lbs. (b) 
 
 Y=:^. 1M80X 23 X23.040. 000 ^ . 
 
 a 27 28,000,006x105.5 =^o.6o inches. (c) 
 
 ^=-^%^=5S.26 tons=106,520 lbs. (d) 
 
 There may be two cases in which a, h, c and d will occur: 
 1st When the load, beginning at one end of the span, covers 
 two-thirds of its length, in which case S is at the sections of 
 the chords at the middle of the loaded segment— the lower 
 chord being in tension and the upper in compression. At 
 same time (c) is the deflection of the middle point of the 
 loaded segment, while (rf) is at the ends of the loaded segment, 
 a, c and d all act downward. 
 
 2d. When the load begins at the end of the span and 
 covers one third of its length, the eflfect upon the unloaded 
 segment is the same as in the first case upon the loaded segment, 
 except (a), (c) and (d) all act upward and (b) is tension on the 
 upper chord and compression on the lower. 
 
 Max. F=:^°=60 tons=120,000 lbs. 
 
 (e) 
 
 e occurs when one-half the span is loaded beginning at one 
 end of the bridge. It acts downward at the extremities of the 
 loaded segment, and upward at the extremities of the un- 
 loaded segment 
 
IV 
 
 
 
 STRAIN Dl'E TO TEMPERATURE. 
 
 At mean temperature iliere is no strain upon the parts of the 
 trusses when unloaded. 
 
 But in winter the cables rising 12 inches lift the trusses 
 the same amount while in summer their own weight bends them 
 downward 12 inches. In both cases it is a question of bending 
 a girder 12 inches by a distributed load of uniform intensity. 
 Then by (7) 
 
 W= 
 
 12 X 4R X 28,000,000 X 1.113.000 
 
 6 X 110,692,000,000 
 
 =32,462 lbs. 
 
 From (8) 
 
 S= 
 
 82,462 X 800 
 
 =3,692 lbs. 
 
 (9) 
 
 (h) 
 
 8X17.58X60 
 
 (g) acts upward in winter, producing a shearing strain which 
 is greatest at the ends of the bridge and equals — 
 
 w_32^_^g 231 lbs. (t) 
 
 (g) acts downward in summer with the same force at the 
 ends of the bridge=^^ lbs.=16,231 .bs. 
 
 (h) acts on the chords at the middle of their length. In 
 winter it is compression on the lower and tension on the upper. 
 
 In summer it acts at the same points but reverses the direction 
 of the strains on the chords. The strain on the chords, due to 
 temperature, diminish frem the middle toward each end as the 
 ordinates of a parabola whose parameter=34,077. Hence from 
 the equation of the parabola. 
 
 y*=2 p X in which x=134 feet. 
 
 y»=34,077xl34=4,656,318, y=2,137 lbs. (k) 
 
 {k) gives the stress, per square inch, on the chords at a point 
 267 feet from the end of the truss, or where the maximum 
 stress from the load comes. 
 
 Examining the preceding values we find as follows : 
 
 1st. That to get the greatest strain per square inch on the 
 chords, which will occur at either extreme of temperature, we 
 add (b) and (k) and have 16,180+2,137=18,317 lbs. (l) 
 
 In winter with the load one-third as long as the bridge 
 beginning at one end, I is at the middle of the unloaded seg- 
 
ment, is tension on the upper chord and compression on the 
 lower. 
 
 In summer, with the load two-thirds as long as the bridge, I 
 is at the middle of the loaded segment, is tension on the lower 
 chord and compression on the upper. 
 
 2d. The greatest shearing force occurs at the ends and is 
 found by adding (e) and (t)=120,000+16,231=136,231 lbs. (m). 
 
 In winter this occurs at one end of the bridge while the load 
 is at the other end and acts upward. 
 
 In summer it is at the end of the loaded segment and acts 
 downward. 
 
 This stress is resisted by six posts, each composed of two 
 6 inch channels, each having a section of 2.1 square inches, or 
 the total section of 6 posts = 25.2 square inches. Consequently 
 the strain per square inch = ^§Ji = 5,406 lbs. 
 
 The stress (m) is also resisted by 12 truss rods whose united 
 cross section = 15 square inches, the tangent of whose incli- 
 nation to the vertical is 1.31. Hence ^^"'''y^'^^ = 11,898 lbs. per 
 square inch. 
 
 The anchorage for holding the ends of the span are secured 
 by eight Lewis bolts, at each end of the bridge. If inches in 
 diameter and sunk into the rock to a depth of 5 feet. 
 
 As before remarked the above calculations suppose a deflec- 
 tion ot 15 inches at the middle of the loaded segment But 
 the formula gives it as over 28 inches which would decrease 
 the intensity still more and hence reduce all the quantities 
 given above. 
 
 The stays will also diminish the strains on the truss members 
 especially those subjected to the shearing force at the ends. 
 
 NOTES ON THE TEST OF THE NEW WORK. 
 
 The actual deflections, under a load consisting of an engine, 
 tender and thirteen box cars loaded, having a total weight of 
 357 tons, are given for eight different positions of the load by 
 Figs. 2 to 9 inclasive (Plate VI). The total length of the train 
 was 465 feet. In the diagrams the vertical scale is 12.5 times 
 as great as the horizontal scale. Fig. 2, Plate V, is constructed 
 from Fig. 5, Plate VI. 
 
i 
 
 111 
 
 I 
 
 '''^ =7.66 tons=16,320 lbs. 
 
 VI 
 
 By applying the parallelogram of forces to the two segments 
 ot the cable (Fig. 2, Plate V), it is found that with the load in 
 the position shown, covering a little over half the span, the 
 intensity of the dead load of the unloaded segment balances 
 not only the intensity of the dead load of the loaded segment, 
 but .12 ton of the intensity of the live load beside. The total 
 intensity of the live load=|5=.77 ton. Consequently {w) in the 
 formula3=.77— 12=.65 ton. Substituting this in (2) we get— 
 
 j^ .65(100 + i5)M^0»-15)_6J33 tons. 
 *• 0*00 
 
 Stram per square in, on chords^ ^^g^^gp 
 
 F=''^°""=;.— '=64.9 tons=129,800 lbs. 
 
 Hence, strain per square inch on posts=5,161 lbs. 
 
 " " " " " truss rods=ll,336 lbs. 
 
 REMARKS. 
 
 The train considered in the above calculation is about the 
 heaviest that will have occasion to cross the bridge and is about 
 as heavy as can be started by an engine on the up grade of the 
 New York side. 
 
 The overfioor stays will afford a very considerable assistance 
 to the trusses, enough at least to compensate for the increased 
 stress due to temperature. An experiment with a Vernier 
 scale, reading to rjshxs foot, applied to the lower chord at a 
 point 200 feet from one end to determine the elongation of 21 
 feet of its length, under a load of one engine and ten loaded 
 box cars weighing about 280 tons, resulted in showing an 
 elongation corresponding to less than 10,000 lbs. per square 
 inch. The stays were not acting at that time. 
 
 The full and dotted lines at ends of Fig. 7, Plate VI coincide 
 very exactly, apparently indicating that with the load in the 
 position shown, the trusses are sufficiently rigid to distribute 
 the load over the cables properly and consequently to produce 
 no more effect upon them than if the load of same total weight 
 was lengthened so as to cover the whole span. Figures 5 and 9, 
 Plate VI, indicate a pretty exact adjustment of stays, sus- 
 penders and truss rods. 
 
Vll 
 
 KEMARKS ON TESTS OF STEEL. 
 
 The tests of specimens of steel for the suspended superstruc- 
 ture were made on the testing machine of Mr. W. L. Gill, 
 manufacturer of car wheels in Allegheny, Pa. 
 
 The strains are given by a screw which enables the operator 
 to stop at any stage of a test and jet be sure of the stress re- 
 maining constant. The strains are indicated by a beam scale. 
 Elongations are measured by means of a michrometer with two 
 screws reading to j^U^ inch. Contact of screws was indicated 
 by a battery being connected and striking a bell. 
 
 It was a very satisfactory machine to work with. 
 
 Enough of the tests are given to show the action of the 
 specimens under severe treatment. (See table of tests.) 
 
 No. 4 was cut from a plate by a planer. The tool raised a 
 " lip " on the corner. In such cases, shortly after the specimen 
 began to stretch, a nick would appear in the " lip " and when 
 the specimen broke it would generally be at that point. It 
 would break by tearing apart. 
 
 Nos. 5, 6, 7 and 8 were all cut from the same plate to test the 
 eflfect of punching and reaming and also of annealing. Regard- 
 ing 5 and 6 it appears a little remarkable that the punched and 
 reamed specimen shows a less elastic limit but a larger ultimate 
 than the plain specime:?.. I account for the less elastic limit by 
 supposing that the stress was greater on one side of the hole 
 than on the other, hence causing stretch to begin on that side 
 first. The cause for the larger ultimate was no doubt due to 
 the hole having the same eflfect that a semicircular groove does 
 in a specimen, viz., by preventing reduction fracture must take 
 place simultaneously over the whole section instead of tearing. 
 
 The effect of annealing appeared to be the same in nearly all 
 cases, viz., to increase not only the stretch and reduction but 
 also the elastic limit and ultimate. 
 
 No. 13 was cut from the same bar as 12. It was nicked on 
 one side with a sharp chisel and when under a strain, consider- 
 ably above the elastic limit, it was struck smartly with a hand 
 hammer on the side opposite the nick, but showed no weakness. 
 In breaking, the break started at the nick and gradually made 
 its way through as in soft iron. 
 
l!W! 
 
 ;;!♦'•' 
 
 
 Vlll 
 
 Nos. 14 and 15 were for the purpose of showing the effect of 
 punching without reaming compared to that of punching and 
 reaming. Calling elastic limit and ultimate of the punched 
 and reamed specimen 100, that of the punched was for elastic 
 limit 98, ultimate 84. 
 
 No. 16 was nicked across one side and after determining the 
 elastic limit it was subjected to a strain of 69,370 lbs. per 
 square inch, and while the strain was on it was struck on the 
 side opposite the nick sufficiently hard to bend it J inch and it 
 retained a bend of ^ inch with the strain on it. 
 
 Nos. 17 and 17 ' were the same specimen. The elastic limit 
 and modulus were first determined. It was then pulled to the 
 maximum after which the strain was removed. The specimen 
 measured for length and cross section, and then treating as a 
 new specimen, it was tested for elastic limit, modulus and 
 ultimate. The reduction given is the total reduction. The 
 modulus is about the same for both tests. The object of this 
 test was to determine if, in case one part of a bar of this grade 
 of steel should receive a greater stress than the other parts, even 
 to causing it to pass the elastic limit, its modulus would still 
 enable it to resist in proportion to its section as much as before. 
 If its modulus is not altered by stretching, the structure of 
 which it formed a part might not be immediately rendered 
 unsafe, but it is more a matter of curiosity than otherwise as 
 nobody would think of straining material to that extent. 
 
 From a study of the tests of steel it appears that below the 
 elastic limit, for a direct tensile stress, a slight nick would not 
 be dangerous, but that the danger from the nick lies more in 
 having transverse vibrations take pla^e which, concentrating a 
 heavy stress at the bottom of the nick, would cause it to break- 
 
 The tensive strain that was on the specimens at the time of 
 striking them was no doubt an assistance in preventing trans* 
 verse vibration in the bar, causing it to resist the effect of the 
 blow, where the same blow would have broken it without the 
 tension. 
 
 The modulus of steel, being so much less than that of iron 
 in proportion to their elastic limitc, gives to the trusses greater 
 flexibility for the same depth of truss when made of steel than 
 if made of iron. L. L. BUCK. 
 
 'y«ff.'!'R='.»»<"--^'Tf»«WWWM?nW'^» 
 
o 
 
 M 
 M 
 
 g 
 
 
 s 
 
 •< 
 
 Q 
 
 i 
 
 o 
 
 K 
 
 P 
 U 
 
 H 
 
 H 
 
 U 
 
 
 o 
 
 o 
 
 X 
 
 CO 
 
 3 
 
 pa 
 e2 
 
 W ^ 
 u 
 
 H 
 V) 
 
 fa 
 
 o 
 
 M 
 
 I*, 
 
 O 
 H 
 
 H 
 O 
 
 u 
 
 o 
 
 c« 
 
 u 
 
 e 
 
 o 
 u 
 
 a 
 o 
 
 1= 
 I. 
 
 cu 
 
 a 
 
 
 u 
 
 ^ s 
 
 V o 
 
 .a J: 
 
 o a 
 
 .a S"* 2 2 e 
 a S S* ^ 8 
 
 B « S M S u 
 3 ►= 5 ^ g « 
 
 O 
 
 £ 
 
 § 
 
 £ 
 
 C4 
 
 
 
 s 
 
 « M i« «• n ^ *• 
 ^ M w 15 ^ V 
 
 ^v.1 
 
 §« 
 
 g&§-^=?? 
 
 §2P5 
 
 0<u 
 
 »o 
 
 GO 
 
 
 ■^ «•*« 
 
 
 "^ Cu ^ a>B 
 9 o o -a^ 
 
 gilgsa 
 
 (2' 
 
 ■Si .21, 
 
 ^ o 
 
 
 5i 
 
 w TjJ OS ©i 
 
 5I> 
 
 ico vt -^ to ui t^ t^ 00 
 
 'OrHiHrHrHf-li-li-l 
 
 SQO'J<© 
 OOMQ 
 
 00 00 t> OS OS 
 
 ^ 
 
 00 00 
 
 U3I>I>^0 
 OCT<NU5 00 
 
 00 nc 00 OS OS 
 
 0«00( 
 
 ®Sfg 
 
 S-* 00 
 ^00 «o 1-ji «eo c o 
 
 i>'os'e<5"«D CO t> TjToo 
 OS 00 OS 00 00 00 t« i> 
 
 S8 
 
 IHOS 
 00 «0 
 
 m 
 
 os" 
 
 U3 
 
 88 
 
 >«o 
 
 s 
 
 I 
 
 oo .oo 
 
 oo -oo 
 
 oo -oo 
 
 ^ C" , l> O 
 
 t>00 "OOOO 
 
 :88S: 
 
 • ooo< 
 
 .OOOi 
 .O O-'J* 1 
 
 'on 
 
 iiOOO 
 
 • «0 -Xl t- 00 o o 
 
 ; ofl o» c<i o» eo eo 
 
 §o 
 o 
 Hi 
 
 t>i>r 
 
 
 § 
 
 4S 
 
 
 ■«ejy 
 
 O BO 
 
 ^ I 
 
 O i^U5< 
 
 CO cc t^ c 
 
 lO 
 
 I OS 
 
 •*OS-* «0<N 
 
 to u3 us iota 
 
 SIOOSOOOQO 
 (N (N ■« CO O «0 U5 
 2fl t»-«JJ^e«iH i-t WM 
 
 t- i-T 00 rH ©if Op' Q ^ 
 ■*lO»0»OlO-^iOlO 
 
 IS 
 
 t-eo 
 
 coco 
 
 to oo 
 
 to 
 
 too 
 
 iOOS 
 
 »o i>«o« 
 
 OS 00 to t» 
 CO CO CO ^ 
 
 to <M (N (N eo iH CO «o 00 eo 
 
 (N (M ffI(N t* CD to to 1^© 
 
 cototoeoeotoioto t-t> 
 
 OS IP 
 
 00 CO 
 
 osiop(Meo 
 -H rH'^ 00 eo 
 
 00 CO C^ CO ^ 
 
 coosososcdt-t«t« 
 «g(Mi-iciiiO"*-^ 
 ^ ''T ^ eo CO t* t* t^ 
 
 ^ 
 
 ^ 
 
 xxxxx 
 
 XXXXXXXX 
 
 00 00 CO 00© 
 (NIMOOCOIO 
 
 eo©©-«*<tot»«>t» 
 
 CD - 1-1 rH O t^H Tj* •^ 
 
 (Nd-^od n 
 
 (N (M (N © (M t- t- t- 
 
 XX 
 
 lO to 
 OS OS 
 
 CD loeo 
 
 O t»lO 
 
 to coco 
 
 X XX 
 
 Jg to?2 
 
 d 
 
 ©OS 
 
 0» 1-1 fH iH »H 
 
 .a § 
 
 
 ©©©©OS 00I>t>O©©©© 
 
 00 00 
 
 ©iH 
 
 S 
 
 iHOICO'^tO © fc« 00 OS O i-« fid CO 
 
 -itto 
 
 © t-t- 
 
 2 
 
 <5 
 
If 
 
 I: 
 
 ■h 
 
 ^ 
 a 
 
 ssi 
 

 > vv 
 
 
 ■t'. 
 
 I 1 
 
 
 
 .^ii^ ,n~) li^ii 
 
 ;.7.T.TT.T.r;iasB3B 
 
 
 
w 
 
 !lWi^ 
 
 
 pi wiiV^sifiJSsji&i.JistJii^^Aji'g^ 
 
 ■ - ■ -3^-<tl^ .. .. - 
 
 
 
 - .^vSKw-siW. 1 
 
 t ,* 
 
 — --■ ~^— -,\» 
 
 
 _ T..-.J' 
 
 ■ ■ 
 
 * "~" "' ■ — ■ ; 
 
 
 ■ : ( 
 
 
 ! 
 
 
 
 ' ■,- . ^ ..' 
 
 
 ■•■ b- 
 
 i 
 
 
 1 
 
 ■ 
 
 1— --.._. 
 
 
 aran 
 
 ' 1 
 
 i'. 
 
 > I 
 
 
 "^^^^-i;: i!i-s^ 
 
 AflAOAIM 
 
 V 
 
 ^.._J 
 
'■pl'j'^www ^ giw 
 
 =1 
 
 .i^ 
 
 f^ 
 
 i 
 
 f 
 
 \]Z 
 
 V. 
 
 :,i 
 
 
 ■yrXc^S^T^Z 
 
 
 //* 
 
 /A 
 
 ..I 
 
 d 
 
 '■^• 
 
 F^ 
 
 
 
 ,/^. 
 
 
 \ 
 
 \ 
 
 
 .'/ 
 
 / 'M 
 
 Ml 
 
 \ 
 
 A,:,.i:, 
 
 ^V-r-i 
 
 ^.^^■.^ b-^-.^-^.^^ If— ^ 
 
 .'iHJUliilSS': 
 
 i ( 
 
 iir^ 
 
 'Si 
 
 vm-i 
 
 
 
 f-f^ 
 
 i^"V,,«»^-.»««<" 
 
 eU MT ii S ' nJS- i 
 
 ~"7iriiririifMr" ''" " '" 
 
■*»a*l^ ' i, ' iWliUi - 
 
 ■^if/jHit Ki»t^mil^i^^,j,i,t,^aiiMiihmiiJS^,AltJSi^iMi.'^ 
 
Plate II. 
 
 SUSPENSION BRIDGE. 
 
 

 H 
 
 111 I 
 
 Pi 
 
 **!***y?r * f" y,yf- ''!?*'fy''?yy'y»q,;fiyii"jijiMyfyifyy>^^ 
 
 
 t^. 
 
 (.i«g y^ !§ 
 
 ..'J.: 
 
 V 
 
 ,4 
 
 i . si 
 
 'A' 
 
 w-f. 
 
 / 
 
 \. 
 
 H 
 
 f 
 
 f.1 
 
 H 
 
 'v v\ 
 
 h4 
 
 t~ia 
 
 ••^■ii'i 4>,-<yicj< i.TT! 
 
 s1i 
 
 
 ' i 
 
 ^ *^ . «*^'«.»-^ ♦ «*■' »--■ 
 
 "wwipi 
 
 WM 
 

 
 
 ^ 
 
 \ 
 
 \\ 
 
 
 .^i 
 
 =6A-j-=r;:Sac; 
 
 B -' ■ n,VL ."« «' i;'._- ';V ';i ;- " ■■ ^j — , 
 
 o ^ o . iif b '^ 
 
 : ..^ A iJ: - 
 
 ^Se'^r' ■'" ' ' ~ 
 
 •'^■r' . r ? -^; ^ "■ ' . ' " ',- 'r' 
 
 " :;.t^r~.:j»^ V 
 
 
 "^^ "' 
 
 
 
 1 
 
 1 
 
 I 
 
 '-;/H^ 
 
 
 
 i-;fc=:-A^~- 
 
 ■■^' 
 
 
 o o o o O ^' 
 
 O D' 'O 
 
 V-. 
 
 o" ■ O' 
 
 ■^ 
 
 o o c • a o 
 
 o 
 
 O O • '3 '~ 
 
 
 6 o do a o 
 
 c' " o' * 
 
' "'f^'-J''^- 
 
 
 
 ■A-',' 
 
 % , 
 
 
 
 
 BaMsaoam 
 
 
 WBKtKtM 
 
 
 '■:'• tf . 
 
 
 ) o 
 
 O 
 
 
 o 
 
 aata 
 
 BBssjl 
 
 Miiii^^aisa^a 
 
 ^^ 
 
mimauwu ' i .".jauwt - i 
 
 •9a!!".'«""~-«r- « i i' " i , iii i ;. i| ,mi. i j. p j ). BwriW1i 
 
 W" 
 
 i 
 
 ■X. 
 
 \ 
 
 -:^fiWaj™Bs.-«T!--({^|^- : 
 
 ^Si'i^ 
 
 V 
 
 
 I' 
 
 ft 
 
 it' ; 
 
 ii. 
 
 nT- 
 
 €■ 
 
 
 Sa-.:-a s;;,-J7~;.-a'»il'<'— T mi-.!gia..!L n--.>,f^ , j.-.-^-',- t.-iv -. slcy?!fi-»T.».t>;-y g;,^ . -, j..- 
 
 1 
 
 i 
 
 I 
 
 'i^«'! 
 
 
 r.aa«sa.i.iaiiaaiagiis 
 
 "1^^ 
 
 „:aSS''^'°'^fflT"fimiiiiiiMyi ii i[|»iiii 
 
Tra-nsvey^e Vtew shfftt^tKf relative /yesUiotis of old ufzd>tr,wt4/crik. 
 
 ■mi 
 
 ■WHi 
 
V M^orjk. 
 
 SidP jaieyation o^ Old (I\^usa. 
 
*i 
 
 
 / 
 
 K^J?^ 
 
 -^ 
 
 '>■ V 
 
 
 --iC --^ 
 
 fe^' 
 
 ^-vl 
 
 
 -J 
 
 :l 
 
 •^1 
 
 a 
 
 i J 
 
 /:■ 
 
 
 
 
 
 
 Nj ■ 4^ 
 
 
 i 
 
 . f 
 
 1 
 
 
 
 
 
 ;#'-l 
 
 
 1 
 
 
 
 
 ■■•"■' ■■^--►i."- 
 
 , 
 
 ■ 1 
 
 
 
 
 ... .wAObCl^-- - . 
 
 *V.^. -. 
 
 
 ' 
 
 *\'-'V^.- 
 
 
 
 • '.-^v^^x^^^ YX: 
 
 
 
 
 
 
 
 
 _:^ — i-^^.;;: -..- r-iTir. irtr, 
 
 
 
 
 spswpr 
 
 
 an«!mfnn!mmm---~r 
 
•TWBBWHf"*!!*- 
 
 «*-*- 
 
 K 
 
 /if 
 
 /v' !] 
 
 ( "01 <:VS 
 
 !.„.: 
 
 OOE 
 
 
 
 
 1 
 
 - •'iti^ 
 
 
 2:,^^ 
 
 '«,AnS. 
 
 -^ .\i 
 
 •V<-'7. 
 
 
 fe^ 
 
 «»-T- 
 
 <«¥*» 
 
 ..&. 
 
 
 ooe 
 
 Kftan 
 
HvLyt-i^ykief — 
 
 A f 
 
 - - - - 60S, - 
 8 0S4-'' 
 
 — — ~60S 8 
 
 TotizZ suspended wei^Tit of i 
 "Weight perfooi rUH. of dead 
 
 - - 9if!, 
 
 
- - - eos, - - - 
 
 8 0S4-''- - 
 
 — -80S 8 - 
 
 spended we-iffftt of hrxdqe^ /OSO tons 
 fr-footr-UH. o/deael load - t.S.9 // 
 
 v^ 
 
 M-i>o 
 
 SOO 600 1*^^ 
 
L 
 
■! i-Jia »' ji, | .ja a u-M"nJi^JL- .lau, iscgea e ' J. ' i-j.ii-L ait ■ n- ua i L«. -»' «Mi « i»»iiJ-iiiij.imujiiiiiiiMii^u » ^ni-iiu,.n uieu i in" f un i ..jvwi i . i -t^r^vii 
 
 .-1 
 
 'O 
 
 A -^A 
 
 itiV-a. -\o Hi 
 
 .-■"o^lHij^Sti^^ ** 
 
 ti«;,t>,\3Ji 
 
 o 
 
 b e 
 
 ? -^ -^J^ 
 
 
 +- 
 
 !•-, 
 
 
 
 ,r 
 
 .v\. 
 
 
 ^ .^.^'A 
 
 
 4.-^*'.r'»y«n»t.-»f^'ni^y^; ■■- ■^^.-t.y! ■-•Aji^ 
 
 o> 
 
 -'■n^^c 
 
 ■..■^ 
 
 ■ tya^^ H -*j.." 
 
 v'i^?.* 
 
 
 — { .-, — 
 
 
 ■-"1 
 
 in 
 
 'a 
 
 -'.'"'" • ?*«^''^i>»'r^i'>>T^i^^J»r°>^^ y L-.*-teAj:«.S^^^^^^''^^^^^' 
 
 OP 
 
 'O 
 
 a 
 
 ^1 
 
 
 
 a? 
 
 13 
 
 — -^ "^tiJ^Mi^lfe^^i ^^i^s:vK--i-<.a»;o -.>^^ 
 
 Of 
 
 5 
 
 Tiilimiff III IIIIWIH HI III IMMM— WB 
 
 
 c? .e .w*^ 
 
 p 
 
 -.:rV'.^,:---^-_,^^,-.; 
 
 *>'««MMH 
 

 » 
 
 
 
 -"^^^^^-^fe^^^^i ^^j ^ ucw L>^ ,^: 
 
 
 
 ?V-7- 
 
 o 
 
 Ok 
 
 d 
 
 
 —^ r7^^-^i>.»4»/ry^.> ^^^>^ S^^oTtfi^^ol^^tf^gUc^oiHak^tf li; 
 
 I 
 
 
 
 -?V5.9. 
 
 oF«et 
 
 d 
 
 » 
 
 100^*^* 200 300 
 
 400 
 
 ■^IS^g^JUJg^ 
 
Fi,. I. 
 
 (S 
 
 A.ctio*tofo 7eL 
 
 ISSssWS^mi 
 
 Plate VI . 
 
 ^US8 yir ttl/Q c^- r-g 
 
 CM 
 
 Ok 
 
 ■[zat<Mj5S'4 f^lfV^ * *^'*" ' 
 
 ^nB^^SyT2r^ff^"Ur^#^ »^f^^'if ^^i/r^^^t»T/a5 a?^ j^ ^OJf^^^^^^^ 
 
 Tig.l. 
 
 o 
 
 
 I 
 E 
 
 i6ooViS*o6)4i2£*>o \i7oS h Wa^oo\u9*foa\^z. 
 
 to 
 
 I 
 
 
 ^ 
 
 5 
 
 CM 
 
 na^m^s— _.-_-k=rt 
 
 /•tf.9. 
 
 I 
 
 i^sa"^:^ 
 
 400 
 
 5 
 
 500 
 
 M 
 
 I 
 
 600- 
 
 «o 
 
 3 
 
 I 
 
 700 
 
 I 
 
 900