Heating of a Disk by rapid Rotation in vacuo. 227 



may be supposed to arise in particles very slightly attached to the 

 disk ; this is radiated off chiefly during rotation, and probably does 

 not greatly affect the disk afterwards. 



(2) There is a surface gas-effect, which in an aluminium and 

 even in an ebonite disk is conducted into the interior as it arises, 

 so that it does not greatly radiate during rotation of the disk (art. 

 26). In a paper disk, however, which is formed of a badly con- 

 ducting material loosely put together, part of the effect does escape 

 as radiation during rotation (art. 27). 



(3) There is a residual effect, which is more deeply seated than 

 the gas-effect. And inasmuch as radiation takes place from a 

 perceptible depth, this effect is much more influential than the gas- 

 effect in increasing radiation after rotation. In the case of a paper 

 disk, this deeply seated effect will be less diminished by radiation 

 during rotation than the gas-effect, and therefore after rotation 

 in such a disk we might expect the gas-effect to be peculiarly small 

 (art 35). 



47. In the course of these experiments we have endeavoured to 

 prove that this residual effect is not caused by vibration. The ra- 

 diation-experiments with aluminium disks of three different thick- 

 nesses went, on the other hand, to show that it was of the nature 

 of a surface-effect. This is confirmed by the results derived from 

 tapping ; for, in the first place, the experiments of art. 45 show 

 that the two effects (the residual and the gas-effect) are probably 

 distributed in the same proportion, going from the centre to the 

 circumference of the disk. Again, taking the two disks of thick- 

 ness -05 and -025 of an inch, we obtain the following results : — 



Effect for ^ hyd. Effect for ^ hyd. +^ air. 



Thin disk 48 (22 observations). 228 (10 observations). 



Thick disk 29 (20 observations). 108 (10 observations). 



Now, allowing for errors of experiment, we see that the resi- 

 dual, as well as the gas-effect, is reduced to about one half for the 

 thick disk. 



Again, an experiment of a similar nature gave the effect for ^ 

 hyd. in an ebonite disk of ^ in. in thickness =33, against a result 

 = 55 for the thin ebonite disk. Unfortunately it was omitted to 

 make a comparison with these two disks for the gas-effect ; never- 

 theless these results are all in favour of the residual effect being a 

 surface-effect. 



48. It might be well to make one remark regarding these experi- 

 ments. They are not like the radiation-experiments, which required 

 an extremely delicate instrument in order to give a sensible effect. 

 But, on the other hand, the effect obtained by tapping being that 

 due to the mere surface of the disk, is liable to be altered by any 

 thing which affects the surface of the disk. We have come to the 

 conclusion that in such experiments it is unadvisable to use a po- 

 rous hygrometric surface, such as that of paper, not having its pores 

 filled with paraffin or some other similar substance. It is likewise 

 desirable that all parts of the apparatus should be as nearly as pos- 

 sible of the same temperature ; indeed we suspect that some ex- 



Q2 



