1 842. J 



THE CIVIL ENGINEKIl AND ARCHITECTS JOURNAL. 



either by welding it lo tlic ciul itself, or by liendiiig over a porlioii of tbc 

 liar, it apiicnrs to iiic very dilticult to prevent diircreiiccs in length of at least 

 a millimetre between tlie bars. 



iNow if there is this difterencc between bars, 5 metres (lG-4 feet) in length, 

 the shortest luust lengthen a luillinietrt:, or "0002 of their length before the 

 others draw. 



lint we know that a tension of two kilogrammes per sqnarc millimetre of 

 scelion produces npon a bar an elongation of 'UOOi of its length. 



The bars then of which we are now speaking, arc strained to the amonnt 

 of I kilogrammes (9 lb.) (per sq. niiHinictrc) before those busiile them sniTer 

 any tension ; what will this amount to if the differences in length arc more 

 than one inillinietre ? 



We see then that tlie problem of equal tension is as difficult of resolution 

 fur a complicated system of chains as for iron cables, for supposing that in 

 the two systems, the excess of tension, either of one wire over another, or of 

 one bar over another, is the same, tliis c.\eess will be a much smaller fraction 

 of the absolute strength of wire than of bar iron ; moreover, the manufacture 

 of cables aliords a greater hope of pcrfectio:i than that of chains.* 



We sec now that the second objection has no raoie weight than the first, 

 lo decide us in favour of wrought iron chains. 



On the other hand, there are objections against the employment of bar 

 iron more dilhcult to remove, and wliich will give additional strength to the 

 reasons which have induced me to yield the preference to iron cables. 



These objections arc as follow : 



1st. The greater part of chain briilges which have fallen have given way 

 at the bolts which unite the links. Now it is extremely ditlieuU to calculate 

 the strength wliieh should he given to them, as wc do not perfectly under- 

 stanil the manner in which they resist the strain ; if we compare them to 

 bars placed upon lixed bearing points, and charged with a weight in the 



middle, and the resistance of which is derived from a formula P = M -r — 



wc find the dimensions very small, too small indeed, according to many 

 experiments. 



If we suppose them to resist as if drawn in the direction of their length, 

 (and many constructors admit this hypothesis,) we arrive at large sections, 

 which greatly increase the weight of the joint ; besides this, wc liave no cer- 

 tain information as to the nuality of the iron w hich they need ; it should not 

 be so soft as that of the chains, because no curvature is required; but still it 

 should not he brittle. To avoid mistakes prejudicial to the durability of the 

 work, it is wise to m.ike them rather too strong than too weak ; hut as I 

 have just said, an increase of weight is the consequence of this precaution. 



•Jnd. The making of the eye rc(iuires great attention ; it has been observed, 

 that when the holt is too Uttle, during the proof, a rent takes place from the 

 outside to the inside of the head of the eye ; when the bolt is too large, the 

 rent opens from the inside toward the outside. Now, as it is almost impos- 

 sible that the work of man should he jjcrfect, in order to avoid the incon- 

 venience above-mentioned, several constructors have proposed to swell out 

 the head of the eye, in order to give it greater strength ; but to do this we 

 must re-disturb the particles of iron by hammering after having heated it, an 

 operation which I have already designated as faulty. 



3rd. During the oscillatory motions, which take place in all suspension 

 briJ;cs, the irons rub forcibly against encli other at all the joints, and this 

 tends to wear them in those parts which have the greatest strain. This 

 inconvenience does not exist in wire cables. 



4th. In the moorings we are compelled to use curved irons, which have of 

 course been re-heated, and are most often squared; this new heating, and 

 the difficulty of proving them, obliges us to give a greater size, which involves 

 another increase of expense. 



5th. In very cold weather iron becomes brittle ; wire enveloped in grease, 



* To make the bars as equal as ]>ossiblc, we can indccil, after having bent 

 and ueldcd the ends, drill llirougli all of them uliich make up u link of the 

 chain, Hhen rnld ; but it is evident in this en.->e, that to prevent the driliin.; 

 from diminishing the strength of the c)e, \ic must either give greater size tii 

 the bar in ibis part or llatti-n it, involving' u lieaiuiK »hieh injures Uic quality 

 of the iron. 



\ There arc, fur instance, some red shear (hot short) irons wblcli lose 

 nothin^ by being wrought at a Mhitc heat, but whirb are injured in qicilily 

 when uriiuglit at a leu rlevalr'l tempcr.tlnre. IJut the wurkni"!! cannot con- 

 fine ihi'inselvcs l<> lullii these ciiwlilions exactly, whence it liippcns, In ir<<iis 

 uheh li.ivc even been proved, that llic rc-forgxl |nirtiiiiis Ixcoine lud, ami 

 cun.scquenlly, that only the fuuil piouf can give evidence of thi.t im|>eXiCctiun ; 

 but OS wc have jus: auid, it is very tliOicuU tv proTe curvcU Uiri, 



and not in immediate contact with the air, must be less brittle than naked 

 bars of iron. 



Cth. Cables in bridges of great span, can be much more ctsily raised to 

 their places than chains. In the proposal for the bridge of Koelic llcriiard, I 

 have calculated that the weight of a cable would be 7,U0K kilogrammes 

 (17.G7.1lb.), while each chain would weigh 31,19G (69,102 11).); a work 

 eipially dilHcult, it may he said, has been executed at the Menai bridge ; but 

 if this proves that it is not impossible, it docs not prove that it is not very 

 difficult. 



Jl. Vicat has asserted, that wires before breaking, suffer a considerable 

 elongation, which announces the rupture beforehand, and thus gives time to 

 make the necessary repairs, while chains break instantaneously. 



This advantage of iron v^ire has been disputed, in the case of wires united 

 in bundles by ligatures, and the interstices of which arc filleil by grease. M. 

 Frinot thinks that these bundles form a brittle system ; he doubtless would 

 like to say, as brittle as bar iron. In suppoit of bis opinion he h.ts cited the 

 fieme, a sort of skein of hempen tlireail, which has most strength when its 

 elements are free, and loses part of it as soon as the loose threads arc bound 

 together, and approximated to the condition of lopcs. If this assertion is 

 confirmed by experiments — and I have prepared some for Ibis puri>ose — 

 cables will in this point of view, be neither worse nor better than bar iron.* 



Third oi/ec/ioii.— Cables form a less rigid system than bars of wrought 

 iron, so that the horizontal vibrations of the roadway arc much greater in 

 the former than in the latter. 



For equal curves and weights this is true, but when wc have once given 

 the preference, even in point of durability, to wire over bar iron, (and I con- 

 fess I have done so,) will we not gain more by increasing the rigidity, by 

 meaus of the greater weight of heavier timbers for the roadway, and by 

 diminishing the curvature of the cables, or the tension which they should 

 bear per square millimetre of section, than by substituting chains for cables .> 

 These latter likewise admit of an arrangement which cannot be adopted 

 for chains. I refer to the cradle form ; and in this case the outside cables 

 being in a plane, inclined from the vertical, have a tendency to draw the 

 whole roadwav towards them, and as this takes place on both sides, it follows 

 that the roadway is kept in its position better than it would be by means 

 of stays. 



1 offer these reflections to the readers of the Annate; as the result of )M;r- 

 fect conviction in my own mind, after deliberate consi.lcratioii, and I can 

 indulge the hope that this conviction will be shared by at least a small 

 number of my associates. I shall examine in a snhscciucnt article the advan- 

 tages and inconveniences of a diminution of curvature, and the defects in the 

 proof wliich chains and cables undergo, cither before or after being pUced. 



Siif;"Cslionfor the Pnienliim n/ Fin- in <hurches.—\ correspondent of Tlir 

 T'lmcsrremarkins upon the frequent occurieiice nf fires in churches, as well 

 as other old buildings, in which the modern practice of warming by stoves 

 has been introduced in buildings not originally coiislmcted for that purpose, 

 suggests that if the nl.in were adopte-.! in all new churches, and in such of the 

 old as are capable of it, of sinking a well near the crcal western di.or of the 

 ehiireh, and fixing a foriini; pump tberetn. lo the spout of which ;i short 

 hoscmight be adopted, and which miKbt also be so contrived .is to fit Uic 

 hose of the parish engine, and be under the caie of the sexton, many very 

 destructive (ires might be prevented by the timely application of water thus 

 easily supplieil on the spot, as in most cases where a lire occurs it is small at 

 first 'brciiking mil, ami a f.w buckets well applied wimld bo sumcunil.i extin- 

 guish it ; while, as is well known, from llie extreme dilTicully which now 

 occurs in proeurinn water at these generaliy isnlite<l buildings, the lire has 

 made an alarmin-.,' progress bcforo any supply of water can Iv nimle available. 

 In some places, the rain water from llio root might be eunductejl into a lank, 

 whicli wouhl answer the purpose c<|ually well, and al-o be made very usrlul 

 when repairs are being made, and lor the purposes ol aljlulion, now so much 

 neglected, from the absence of water in the churchyard. 



• M. Lauraut recently atlended to ihe work of the bridge of Roche Ber- 

 nard, and has given me some details in regard to the *rr«-, whicli can explain 

 the objection of M. Frinot. The /.rr...- is made by laying the threads in a 

 circle, and if they ar,> perfectly parillcl and evenly stretched, the inside ones 

 arc shorter than those ouuide ; if Hgalures are appliel. and the hfif u 

 drawn out, as it must be to lift or pull loads, wc see that the thrc.iils of the 

 bundle, which Iron, being circular are now slraigbl. or even bent in opiwaitc 

 directions-thc outside ihrc;ids Uien must be thrust up logcllier, while the 

 inner ones, if Uiey do nut stretch, must break. Without the hgaturos Uiere 

 would be a rumiK'noiioii l.twc^n this Icngllicning and »li(.rlening, so that 

 they actually injure the strength of Ihe bundle ; but it is llie rh.inge of form, 

 and' not the ligature, whirh is the cause of the evil, and the hrrir, b<mnd or 

 not bound, wuukl have the s;uiie slicnBll'. "f U ictmncd its primitive form, a« 

 IS the casv Willi cables uf wu«. 



