1887.] on the Boiling Contact of Bodies. 131 



as when raotiou is transmitted from the edge of one rotating disk to 

 another, and then this distinction still more closely holds, as the 

 normal jjressure is only obtained so as to insure the necessary tangen- 

 tial resistance. Thus the objects of rolling motion may be classed as 

 being — 



(1) To allow the relative motion of one body to another with 

 which it is in contact without permitting relative motion of that part 

 of their surfaces in actual contact. 



(2) To ohiain the relative motion of such parts of the surfaces of 

 bodies as are not in contact by means of statical contact of the j^arts 

 which are. 



The lecturer then proceeded to consider the practical proofs of the 

 smallness of the resistance to rolling in cases where the distortion of 

 the surfaces in contact is very small, as illustrated by the small trac- 

 tive force required for heavy bodies properly mounted on wheels or 

 on roller-bearings ; mentioning the case of a 12-horse-i30wer engine, 

 the shaft of which continued to rotate for three-quarters of an hour 

 after the motive power was withdrawn ; and another case, of a turn- 

 table weighing 14 tons, whicih was kept in motion by a weight of 

 SJ pounds acting upon it by means of a cord passing over a pulley. 

 The small distortion of such surfaces when transmitting motion 

 requiring expenditure of energy to maintain, was next made clear by 

 giving certain facts as to the accuracy with which one surface was 

 developed or measured out upon another. An account was given of 

 experiments made with apparatus specially prepared by the lecturer 

 to investigate this point. Tliis apparatus consisted of two accurately 

 turned brass disks pro^Derly mounted upon a frame, and the relative 

 positions of these disks could be interchanged so that any minute 

 differences in their peripheries could be detected. The experiments, 

 which were very difficult to carry out accurately, showed that under 

 the best circumstances, motion with an error of only 1 in 300,000 of 

 the distance passed over could be obtained. This accurate measuring 

 out of the surfaces one upon another was employed in various ways 

 for purposes of measurement, and these, by means of models and 

 diagrams, were briefly explained. 



Although the foregoing facts prove that, under suitable conditions, 

 distortion at the points of contact is very small, yet some resistance 

 at these points always occurs, because no bodies are perfectly hard ; 

 and the nature of this distortion and consequent resistance was next 

 discussed. 



The explanation of the resistance opposed by a soft surface to a 

 bard body rolling upon it, as first given by Prof. Osborne Eeynolds, 

 was applied by the lecturer to account for a very remarkable effect 

 produced in the disk, globe, and cylinder integrator of Prof. James 

 Thomson. This effect, which was the turning of the cylinder when 

 the sphere was rolled along it in a horizontal direction, was repro- 

 duced by means of a large model. The action of a soft body rolling 

 upon a hard surface was next considered, with the result of showing 



K 2 



