April 24, 1885.] 



SCIENCE, 



345 



That the elastic yielding of the ballast under the 

 passing loads, and the slight rocking of the ties,. ab- 

 sorb or resist the creeping force, would appear from 

 the fact that the tendency to creep is most pro- 

 nounced where the supports under the rails are held 

 rigidly, as in bridges. On the Harrisburg bridge, 

 over the Susquehanna, the Pennsylvania company 

 encountered this difficulty, but arrested the move- 

 ment by spikes through the angle-splices at joints. 

 On the St. Louis arched bridge, and its east approach, 

 there is found a most remarkable example of creep- 

 ing rails. Prof. J. B. Johnson, in a paper read before 

 the Engineers' club of St. Louis, 1 discusses this case 

 at length, and offers an explanation. 



The bridge proper is 1,600 feet long; the east ap- 

 proach, a series of short girders on iron columns, is 

 2,503 feet long, with a grade rising towards the bridge 

 of eighty feet per mile; both are double-tracked. 

 As it was thought by those in charge of the bridge 

 that fastenings at frequent intervals, to resist the 

 movement, would bring too great a strain upon the 

 structure, the attempt was made to restrain the rails 

 by holding them firmly at isolated points some dis- 

 tance apart, with the result that spikes, bolts, and 

 splice-bars were sheared off or torn apart. After the 

 failure of attempts to arrest the creeping, the track 

 was cut at the two abutments and at the east end 

 of the east approach. The time of eight men (five 

 by day, and three by night) is stated to be largely 

 occupied in changing rails at these points. Where 

 the openings are enlarging, short pieces of rail are 

 taken out, and longer ones put in their place : where 

 the openings are closing up, the process is reversed. 

 Each operation is performed many times a day, and 

 a careful record is kept, from which the following 

 facts were obtained: the north track, when carry- 

 ing an annual westward traffic of about 5,283,000 

 tons, moved west on the approach and up-grade 401 

 feet in a year, and on the bridge moved 264 feet; the 

 south track, under an eastward traffic of 4,807,000 

 tons, crept east 414 feet on the approach, and 240 

 feet on the bridge, in the same time. The movement 

 each way on the bridge was proportional to the ton- 

 nage ; and the difference on the approach was doubt- 

 less due to the grade, as the changes of temperature 

 would produce a slipping down hill, as previously 

 stated. 



Professor Johnson cites some explanations of this 

 case that have been given: viz., the stopping of 

 trains on the bridge; the deflection of the bridge 

 itself by the weight of the train ; the distortion of the 

 arch, as a train enters a span, by its curve becoming 

 less convex on the loaded portion, and more convex 

 on the unloaded side, with a reversal of the distor- 

 tion as the train passes over and off the span, the 

 arch thus slipping under the rails ; and, finally, the 

 elastic rolling-out and recovery of the rails under 

 successive wheels, as we may imagine a strip of rub- 

 ber to move as a roller is passed over it. He does 

 not think, however, that these causes are sufficient 

 to account for so great a movement, and, in explain- 



1 Journal of the Association of engineering societies, Novem- 

 ber, 1884. 



ing his theory, offers a preliminary illustration. Sup- 

 pose a span of a bridge to have supports exactly 

 alike, such as sliding surfaces, at the ends of the 

 bottom chord, and a train to enter upon it. The 

 bottom chord is stretched by the action of the load, 

 and, as the end where the engine enters is held fast 

 by the added weight, the other end must slip on its 

 support in the direction of the train movement. As 

 the cars pass off at this latter end, and hold it fast, 

 the lower chord shortens, and recovers itself at the 

 first bearing by slipping towards the train. Thus the 

 bridge creeps in the direction of the moving train. 

 If the points of support were under the upper chord, 

 the direction of this creeping would be reversed. 

 When rollers are placed under one end, and the other 

 is anchored fast, the slip and recovery take place on 

 the rollers, and no creeping results. 



He notes that between the trucks of every car the 

 rail springs up from the support an appreciable dis- 

 tance, by reason of the elasticity of its bearings, and 

 that, when pressed down by the passage of the rear 

 truck, any marked point on it has advanced a small 

 distance. A wave-motion of the rail may be per- 

 ceived in advance of every wheel, and an increment 

 of forward movement every time a wheel passes. 

 The more cars, the more movement for any train. 

 The rail moves across the bridge by reason of the 

 extension under flexure of the flange on which it 

 rests. In proof of his position, he showed, by a model 

 over which a loaded wheel was rolled, that a rail 

 supported by the bottom flange will creep forwards, 

 and that the same rail, when supported by its head, 

 will creep backwards; and hence he argues that 

 some point of support between the head and the 

 bottom flange may be found, for which the tendency 

 to creep shall be zero. 



THE PATRIARCHAL THEORY. 



In. 1861, Sir Henry Maine's work on ' An- 

 cient law ' was published. In that work he 

 clearly set forth the importance of 'legal 

 fictions ' in the development of institutions. 

 In this respect, his work will remain as a per- 

 manent contribution to the science of society. 

 In the same treatise he made an exposition of 

 the patriarchal theory of the origin of society, 

 which had long been held by a class of writers 

 in Europe, hi his introduction he says, — 



" This evidence establishes that view of the 

 race which is known as the patriarchal theorj-. 

 This theoiy is based on the scriptural historj^ 

 of the Hebrew patriarchs. All known societies 

 were originally organized on this model. The 

 eldest male parent is absolutely supreme in his 

 household. His dominion extends to life and 

 death, and is as unqualified over his children 



The patriarchal theory. Based on the papers of the late 

 John Ferguson McLennan. Edited and completed by Donald 

 McLennan. London, Macmillan, 1885. 16+365 p. 8°. 



