598 



NATURE 



[Oct. 19, 18S2 



which opens ; this arrangement gives little enough light and no 

 air. Can it be healthy ? Ought it to be ? It is at least most 

 disagreeable. A. H. 



October 15 



ON THE PROPOSED FORTH BRIDGE 



AN interesting account of the plan of the railway 

 bridge for crossing the Forth at Queensferry, as 

 designed by our distinguished engineer, Mr. Fowler, with 

 the association of Mr. Baker, was given by Mr. Baker to 

 the British Association at their late meeting at South- 

 ampton. Supported as it was, to the advantage of those 

 present, by the exhibition of the model of the proposed 

 bridge, it must have given extensive information on the 

 character of the structure. Yet it seems to me that, 

 amidst many valuable particulars, on the strength of 

 materials, their mode of application in this instance, and 

 similar important subjects — it would hardly impress 

 sufficiently, upon the minds of hearers or readers, the 

 vastness of the scheme, the novelty of its arrangements, 

 and the dangers (yet untried) to which, conjecturally, it 

 may be subject. I have thought therefore that I might, 

 without impropriety, offer to the editor of Nature some 

 remarks on points which after careful consideration have 

 suggested themselves to me. For some particulars I am 

 indebted to the courtesy of Mr. Fowler himself, and I 

 greatly value this kindness. 



It is known that at Queensferry the separation of the 

 river-banks, or rather that of the piers next to the banks, 

 at the elevation required for the railway, approaches to a 

 mile. This space is divided by three piers (for which 

 there are excellent foundations on rock and hard clay) 

 into four parts, but only the two middle parts concern us 

 now. They are exactly similar, and are treated in exactly 

 the same way ; and subsequent allusions, referring 

 ostensibly to one, are to be considered as applicable to 

 both. Each of the three piers is an iron frame, 350 feet 

 high, the central pier 270 feet wide (in the direction of 

 length of the bridge), and each of the others 150 feet. 

 These lofty frames are braced, each upper angle on one 

 side to lower angle on the other side, with no other 

 diagonal bracing, but with a simple tie at mid-height. 

 The lengths of the diagonal bracing are respectively 

 about 430 and 360 feet. The water-spaces between two 

 piers are each about 1700 feet ; and the engineering ques- 

 tion now is, how this space of 1700 feet (roughly one- 

 third of a mile) is to be bridged for the passage of a 

 railway. 



The plan proposed is, to attach to each side of each 

 frame (that is, to each side which will face a traveller 

 entering upon the bridge) a framed cantilever or bracket 

 about 675 feet long (that is, exceeding in length an 

 English furlong by 15 feet), attached at top and bottom 

 to the iron frame above mentioned, but having no other 

 support in its entire length of 675 feet. To give the 

 reader a practical idea of the length of "this bracket, I 

 remark that the length of St. Paul's Cathedral, outside to 

 outside, is exactly 500 feet ; and thus this bracket, which 

 is to project over the water without any support whatever, 

 is longer than the Cathedral by 175 feet. This in itself is 

 enough to excite some fear, supposing the bracket to 

 support merely its own weight. But further, the bracket 

 bears also the very considerable weight of the roadway 

 and rails. It is also heavily loaded on its point. The 

 two opposing brackets from the two iron frames cover 

 1350 feet, but the whole space to be covered is 1700 feet, 

 leaving 350 feet yet to be supplied for the support of the 

 railway. To furnish this, a lattice-girder carrying a rail- 

 way is provided, rather more than 350 feet long, whose 

 extremities rest upon the tips of the two brackets. 



This statement is enough, I think, to justify great 

 alarm. No specimen, I believe, exists of any cantilever 

 protruding to a length comparable, even in a low degree, 



to the enormous brackets proposed here. The only 

 structures of this class, in ordinary mechanics, known to 

 me, are the swing-bridges for crossing dock-entrances, and 

 the like, and these are absolutely petty in the present 

 comparison. 



I now advert to the weights of the principal portions of 

 the bridge, and the strains which they will create. I 

 understand that the weight of the two parallel braced sides 

 of one bracket is about 3360 tons, to which is to be added 

 the weight of roadway and rails for 675 feet, on which I 

 have no information. I proceed to inquire what strains, 

 in the nature of horizontal pull at the top of the pier and 

 horizontal push at the bottom of the pier, will be caused 

 by this weight. If the weight were evenly dispersed over 

 the triangular bracket, its centre of gravity would be 

 distant from the pier by one-third of the distance of the 

 point from the pier. But as no vertical bar near the pier 

 is included in the' weights above, I must take a larger 

 factor, say \. The vertical weight being 3360 tons, acting 

 at a distance from the pier of \ X 675 feet, and the sepa- 

 ration of the points of connection with the pier being 350 

 feet, it is easily seen that the horizontal pull at the top 

 and push at the bottom are each about 2600 tons. The 

 inclined tension along the great upper bar of the canti- 

 lever and the inclined thrust along the great lower bar of 

 the cantilever are therefore each about 2670 tons. The 

 extremities of the great upper bar and the great lower bar 

 being connected at the point of the bracket, and (for a 

 moment) no other weight being supposed to act, there is 

 no tension or thrust at that point, and therefore the ten- 

 sion and the thrust increase gradually, according to the 

 attachment of their loads, from nothing at the point of 

 the bracket to 2670 tons at connection with the pier. 



But the point of the bracket is permanently loaded with 

 half the weight of the intermediate 3<;o-feet railway, or 

 363 tons, and occasionally loaded with the whole weight 

 of a railway train, say for a passenger train 150 tons (a 

 mineral train would be heavier). The vertical weight of 

 513 tons thus introduced would be met by a tension of 

 1004 tons through the whole length of the great upper bar, 

 and a thrust of 1004 tons through the whole length of the 

 great lower bar. Thus we have — 



For the great upper bar, a tension increasing from 1004 

 tons near its point, to 3674 tons near the pier. 



For the great lower bar, a thrust increasing from 1004 

 tons near its point, to 3674 tons near the pier. 



The second of these statements particularly requires 

 attention. 



Mechanical students and professional engineers are 

 accustomed to estimate by numerical measure the magni- 

 tude of a horizontal or nearly horizontal thrust, but 

 persons in ordinary life scarcely attach a clear meaning 

 to such a phrase. I am therefore compelled to make a 

 somewhat violent explanatory supposition, with the hope 

 that it may convey a practical impression as to the 

 meaning of the statements just given. 



The great lower bar is in fact a nearly flat frame, 

 braced from side to side, about 120 feet wide at the 

 bottom, and about 40 feet wide at the top, and 690 feet 

 long. Suppose this structure to be planted vertically, say 

 in St. Paul's Churchyard, without any bars, chains, or 

 any thing else, below its vertex, to prevent motion edge- 

 wise, but with bracing (which, under ordinary circum- 

 stances, would suffice, but which will be the subject ot 

 further remark) to prevent its moving flatwise. Its top 

 would be 310 feet higher than the top of the cross of St. 

 Paul's Cathedral. Suppose a weight of 1000 tons to be 

 placed on its very top, and additional weights (if neces- 

 sary) to be placed at its sides, till the whole weight 

 pressing the ground is 3600 tons. In this state its con- 

 dition is exactly that of the great lower bar, as regards 

 the crushing and distorting tendency of the weights 

 (although the upper weight itself ought to be considered 

 as partially protected from lateral movement by the great 



