424 
-work done per unit of mass is therefore everywhere the same, 
and the temperature ought therefore to be uniformly raised all 
over the surface. This would imply that the gas-effect was 
uniform all over the disk, which is not probable. The explana- 
tion therefore does not seem to agree with experiment here, but 
unfortunately the experiment on which the law is based is 
rather unsatisfactory. In the first place, the temperatures were 
compared at two points only and distant from one another only 
t radius; and secondly, the amount itself to be measured was 
very small. 
3. The quantity of heat taken in would probably be the 
same for different thicknesses of the same kind of disk, for 40” 
would not be long enough for even the thinnest disk to rise 
exactly to the surrounding temperature. 
4. This would certainly be the case, though searceke pro- 
bable on the supposition of etherial friction. 
5. It is clearly independent of the residual gas. 
6. And it is evidently different for different disks. In fact, 
for india rubber it ought to give a cold effect. 
If we consider what the etherial friction would be, it seems 
more probably due to a shearing force, separating the ether in 
the body from the ether in space than a true friction in the 
ordinary sense. If this were so, it would probably be to a great 
degree independent of the material of the disk ; but still it is 
clear that different materials would be differently affected, 
though the effect might not depend on the polish of the sur- 
face. 
The question could at once be settled by the following 
experiments, 
1. There ought to be at first, when the disk is in motion, a 
temporary cooling effect. (The heating effect was observed only 
when the disks were at rest.) 
2. The work done at any point is. proportional to the 
square of the angular velocity, while for friction it would 
