358 



THE CIVIL ENGINEER AND ARCHITECTS JOURNAL. 



[Nov. 



MuIliDS show thai Snieaton and liis successors have entertained erroneous 

 views respecting works cuuslrueted in bog lands, and suggests methods 

 which are recommended by experience and common sense. In addition to 

 the valuable information which it contains, the work has the advantage of 

 being written in a concise and perspicuous style. 



Elements of Physics. By C. F. Peschel ; translated from the German 

 by E. West. Part II. Imponderable Bodies. London : Longman, 1846. 

 I2mo. Vols. 2 and 3. Moodcuts. 



The publication of the second part of this work has been hastened by 

 the rapid sale of the first. This circumstance is a gratifying indication 

 that the public taste for sound knowledge of physical science is on the 

 increase. The volumes before us are similar in plan to that already re- 

 Tiewed {atite Nov. 1845), and treat of light, heat, magnetism, electricity, 

 electro-magnetism, and magneto-electricity. Of all elementary treatises 

 on t!ie philosophy of what are termed " imponderable bodies" which we 

 have examined, the present gives by far the clearest and most accurate 

 account of the state of scientific knowledge up to the present time. The 

 author, from national prejudices, occasionally assigns too large a share of 

 the merit of important discoveries to the experimentalists of Germany ; 

 but even this defect has its compensating advantage, for it enables the 

 ^English reader to regard science from a new point of view, and familiar- 

 ises him with worthy names with which he has hitherto been little ac- 

 quainted. 



The tables of the tension of steam do not seem happily chosen. 

 Southern's formula is accurate for a low pressure only, and TreJgold's 

 is not trustworthy, except for pressures ranging between 1 and 4 atmos- 

 pheres. The experiments of the French Academy apply only to very high 

 pressures. The translator would, we think, have been justified in giving 

 the far more complete tables in De Pambour's treatise on locomotive en- 

 gines. The theory of the power of steam engines, developed in the same 

 treatise, might have been substituted for the totally erroneous theory given 

 in ^421 of the work before us, in which the evaporative power of the 

 boiler is not taken as a numerical ingredient of calculation. 



The number of errors is, however, exceedingly small in reference to the 

 extent and variety of the subjects embraced. The author has, in many 

 cases, given practical hints for the construction of apparatus and perform- 

 ance of experiments, deduced from his own experience : the collection of 

 the matter contained in this treatise must have cost him much labour and 

 research, and his efforts have been well seconded by the translator, who 

 has produced a very perspicuous version of the original texts. 



A Reply to some Observations in a Rerieiv of the Pamphlet entitled Me- 

 tropolitan Bridges and Westminster Improvements, in the Civil Engineer 

 and Architect's Journal, September, 1816. Addressed to the Editor of that 

 Journal. 



Sir Howard Douglas has written a reply to that part of our review of 

 his pamphlet entitled " Metropolitan Bridges, &c.,"in which the accuracy 

 of some of his views respecting the stability of Hungerford Bridge is 

 questioned. The following extracts from the " Keply'' will, we trust, 

 fairly represent the arguments brought against us : we had prepared some 

 remarks in answer to these arguments, but are compelled to defer them 

 till next month. The reasons for difl'ering from Sir Howard Douglas were 

 expressed with caution, and remain unchanged ; but we have to express 

 our acknowledgments of the very corteous spirit in which he has addressed 

 to us the present vindications of his opinions : — 



The reviewer's first objection is, that it is not confirmed by very conclu- 

 sive reasoning, that the strength of the suspension chains ought to be 

 greater than three times the strain to which they may be exposed. This 

 can only be derived from experience. In rigid or inflexible works, this 

 strength may be sufficient; but in suspension bridges, sutiject to consi- 

 derable motion, and an increased strain, arising from vibration, undulation, 

 and momentum, it is quite clear that additional strength ought to be given 

 to those parts which support the strains; and Mr. Davies Gilbert's rule, 

 that the strain ought not to be greater than one-sixth of the weight which 

 the chains are capable of sustaining, appears to be well founded 



The reviewer's next objection is, that the curve is not, as assumed, a 

 common (simple) catenary. This observation is correct, but no error arises 

 from considering the curve as such. In fact, it has been demonstrated by 

 mathematicians that, when the abscissa or height of the curve bears no 

 greater proportion to its length than that which exists in any suspension 

 bridge which lias yet been constructed, the horizontal tensions, or strains, 

 whether the curve be a simple or a loaded catenary (a slender chain, or 



one from which, for example, a heavy bridge is suspended) will varv di- 

 rectly as the ordinate or span, and the length of the curve, and inverselv 

 as the abscissa or height : it may here be added that, when the height is 

 small, compared with tin; span, the curve coincides very nearly with the 

 common parabola (and such curve Galileo supposed it to be). In the 

 pamphlet reviewed, the curve was assumed to be the simple catenary 

 merely because precise values could on that supposition be given to the 

 tensions. But whether a catenary be simple, or whether it be considered 

 as such a curve, modified by the weight of a road-way and suspension 

 rods, the values of the horizontal tensions (corresponding to a or n') vary 

 directly with the length of the chain above a given point, and with the 

 span, or ordinate; therefore the heights, or abscissie, being equal, the term 

 corresponding to a' is less than that which corresponds to a, in a higher 

 ratio than y' is less tlian y, which was all that it was intended to express. 



The reviewer's third observation is, that, in finding the strains on the 

 curves at the two extremities, " a furiiuila has been applied to a case with 

 which it has no connection :" it is added that " the value of a' is taken from 

 the tension in a chain at its lowest point, when the lowest point is hori- 

 zontal ;" and it is observed that " the shorter chain, where it is attached to 

 the abutment, is inclined to the horizon at a considerable angle," implying 

 that there is a difference between the horizontal strains in different parts 

 of the length of a chain suspended between two points. In this the re- 

 viewer has, however, overlooked the first principles of mechanics, respect- 

 ing the properties of the catenarian curve ; and it may be added, the equi- 

 librium of an ordinary arch, since it is demonstrated by mathematicians 

 that, both in the catenary and the equilibrated arch, the tension, estimated 

 horizontally, at every point in the curve is the same. Hence, if a chain 

 suspended from two fixed points in a horizontal line be in equilibrio, and if, 

 while the chain remains fixed at one of the points, any poriion ' e removed, 

 the lower extremity of the portion which remains being attacbiJ to a fixed 

 point, that portion will still be in equilibrio. The abutment chains of a 

 suspension bridge, instead of descending at a considerable angle with the 

 horizon to a point of the abutment much below the platform, as in Hunger- 

 ford Bridge, should be fixed at points so situated that each of those chains 

 may assume a figure precisely equal and similar to half the chain betweea 

 the piers. In Hungerford Bridge, " the chain attached to each of the abut- 

 ynents being shorter than half the chai7i between the piers,"" and '• descending 

 to the groiaid at a considerable angle," is precisely the cause that the hori- 

 zontal tensions of these chains are less than those of the chains between the 

 piers, and that a power (represented by (« — a') is constantly acting at the 

 head of each, to draw it towards the middle of the river. 



The Reviewer appears to doubt this, observing that " if the saddle were 

 acted upon by an accelerating (qy. moving) force as o — a', it would be set 

 in motion;" and he probably intends to imply that the motion of the saddle 

 would prevent the pier from being disturbed. He asserts, moreover, that 

 " the pressure of the rollers upon the top of the pier is normal to the sur- 

 faces in contact, and is therefore irholly vertical." Now, with respect to 

 this assertion, the use of the rollers is to permit the saddles, to which the 

 chains are attached, to shift their places, or slide, on the piers with the 

 vibrations of the bridge; and in so far, certainly, they diminish the shaking 

 of the piers ; but, though the shaking is diminished, it is not removed ; it 

 exists to a great degree on account of the enormous friction, arising from 

 the weight of the bridge, which takes place both on the saddles and piers, 

 and when, in consequence of great strains arising from sudden accessions 

 of weight on the bridge, the saddles are drawn to the limits of their motion 

 on the tops of the piers, they cease to relieve the latter from the eflects of 

 such strains. This is a defect which exists in all suspension bridges ; and 

 even the ingenious expedient of the shifting saddles is incapable of remov- 

 ing it. The evil is, moreover, rather augmented in Hungerford Bridge, ia 

 consequence of the strains introduced by the convex figure which has been 

 given to the platform, the height of the convexity being about four feet in 

 the middle of the bridge. 



AVilh the suspension bridge, if A B and A C, — or to speak more mathe- 

 matically, if the tangents to the curves, formed by the chains at the heads 

 of the piers, do not make equal angles with the vertical line A V, even if 

 the tensions of A B, A C, were equal (which is not the case in Hungerford 



Bridge), the resultant of the strains at A would be in the direction of a 

 line A D, drawn from the point at which the tangents meet, to bisect the 

 angle B A C, and would therefore not be vertical ; thus there is an excess 

 of force acting horizontally towards the middle of the bridge, at the end 

 of a lever whose length is equal to the height of the pier, by which the 

 pier is continually strained, and the uniformily of pressure, on the founda- 

 tions of the piers, destroyed. A very considerable error of this descrip- 

 tion exists in the Hungerford Bridge, " by the shorter chains descending 



