TRANSACTIONS OF SECTION G. 731 



practicable to avoid the formation of these inconvenient undulations. From time 

 to time the crests, which are said to become very hard and compact, are hacked 

 up, and the material thrown into the troughs. Their formation is generally 

 ascribed by those who use the road to initial inequalities (such as cross-channels 

 caused by violent rains) of sufficient size to make the sledge pitch. Unequal 

 hardness of different parts of the road has also been assigned as a cause for com- 

 mencing that pitching motion which is obviously associated with the formation of 

 the undulations. 



When these conditions exist they undoubtedly tend to produce the effect ; but 

 the author has found, by experiment with a miniature sledge, that, when the 

 detritus of the road consolidates readily under pressure, these undulations arise 

 spontaneously by the action of a steadily moving sledge when furrowing a 

 homogeneous and level road. The wedge of detritus pushed along in front of the 

 prow of the sledge becomes compacted, the sledge surmounts it, rolling like a 

 wheel, and the detritus remains behind, as an excrescence incorporated with the 

 road. At the same time the sledge pitches, furrowing the road more deeply, 

 accumulating detritus in front of it, which it finally surmounts (with the rolling 

 movement which assists to compress, bind, and consolidate the material) building 

 up the next crest. 



The conversion of the plane surface to a strongly undulating surface of definite 

 wave length, in the author's experiments about 3i times the length of the sledge, 

 proceeded very rapidly when the material (coarse river sand by the shore of 

 Coniston Water) was of the right degree of dampness. The (horizontal) track was 

 free from inequalities or obstructions, and the sledge was drawn steadily and 

 slowly. 



Corresponding undulations sometimes produced by sledge-driving on snow- 

 covered roads are familiar in Canada, where they are called cahots (jolts). The 

 author saw these in and near Montreal in the winter 1900-1901. They averaged 

 13 feet in length from crest to crest, and 8 inches appeared to be a not uncommon 

 amplitude. The length of the sledge-runners varied, but 5 to 6 feet seemed to 

 be a common size. The author came to the conclusion that for the marked 

 development of cahots the snow should be in the state to bind under the pressure 

 of the sledge when it rolls over a prominence. The dry snow in zero weather at 

 Winnipeg did not show the same tendency to go into cahots. 



The author has investigated many cases of undulating surfaces spontaneously 

 produced by the passage of a Jluid (air or water) over granular materials. The 

 above observations appear to extend the range of this class of phenomena. 



4. A New Elastic Vehicle Wheel. By J. Brown, F.R.S. 



In previously designed vehicle wheels having springs inserted between hub 

 and rim, such springs, if rigidly fixed at their ends, are soon broken off at these 

 points of attachment owing to the bending due to the continual displacement 

 of the hub relatively to the rim during the revolution of the wheel. 



If springs be attached through jointed connections to rim and hub, the action 

 of such joints involves continual loss of power by friction in the joints as the 

 wheel revolves. 



In the wheel shown no joints are used, and the spring spokes are attached 

 rigidly to hub and rim, but the tendency to fracture at the points of attachment 

 is overcome in spreading the breaking stress over a foot or more of the ends of the 

 spring by giving the ends of the steel strip from which the spring is made a 

 taper form, gradually increasing in thickness towards the ends. The spring spokes 

 are crossed alternately between hub and rim to give sideway stability. The outer 

 ends are bolted to a casting attached to inner side of rim ; the inner ends to a 

 specially made hub so formed that by means of a taper key passing through a slot 

 in the tail of the spring its tension can be adjusted so as to keep the rim true. 

 The whole set of springs are then clipped fast by rings bolted on the faces of the 

 hub. The wheel stood well a test at considerable speed on rough roads as a 

 driver, and carrying a load of 11 cwt. 



