308 
MR. W. CROOKES OK REPULSION RESULTING FROM RADIATION. 
quenched by the rapid movement of the vanes in the opposite direction. As is already 
shown/* the viscosity of air at a rarefaction of 129 millionths of an atmosphere is only 
a little less than its viscosity at the normal density ; hence the vanes, at a speed of 
100 revolutions a minute, must exert a considerable drag on the opposite rotation 
of the disk. 
353. Pressure 82'5 millionths of an atmosphere. The disk has again increased in 
speed, now revolving at the rate of 24 turns a minute. The speed of the vanes is 
too great to count ; a rough estimate shows that the rate is about 600 revolutions a 
minute. The direction of the vanes is positive, whilst the disk rotates in the opposite 
direction . 
The increased speed of the disk, in spite of the greater rapidity of the vanes tending 
to drag it round in the opposite direction, shows that the viscosity of the residual gas 
is diminished. On referring to my preliminary note,t it will be seen that the viscosity 
of air has commenced to diminish rapidly at 82 millionths of an atmosphere, and 
after that the fall is very quick. To carry these experiments to a much higher exhaus- 
tion, some modification in the apparatus is needed ; the heat must be less concen- 
trated, so as to diminish the speed, since it is useless to continue observations 
when the rate of rotation cannot be estimated. The following apparatus was accord- 
ingly fitted up, to enable me to get the observation of speed, together with the 
viscosity of the internal air. 
As the results obtained with this new apparatus are more numerous than with 
the previous apparatus, I will for the present defer comments on the different motions 
of the disk and vanes. 
354. The apparatus is the most complicated I have yet used, and in order to 
make clear the relative bearing of the different observations I will describe it in detail, 
and give a drawing of it (fig. 32). This will serve a double purpose; it will enable 
my description to be clearly understood, and it will illustrate, better than is possible 
by mere words, the great complexity of apparatus frequently requisite to obtain, in 
a physical investigation, results which can be expressed in a few short paragraphs. 
355. The pump is shown at a b, the upper part only being represented. It does 
not differ materially from the form of pump described by Mr. C. H. Gimingham in 
the ‘Proceedings of the Boyal Society,’ No. 176, 1876. It has, however, five fall 
tubes instead of three, and is fitted with a small radiometer, c, and a McLeod 
measuring apparatus, d e, to enable me to ascertain the degree of exhaustion in the 
apparatus. The facility of working has been considerably improved by the introduc- 
tion of phosphoric anhydride \ instead of sulphuric acid for absorbing aqueous vapour. 
The phosphoric anhydride dries gases better than sulphuric acid does, and evolves no 
vapour, whereas at the highest exhaustions sulphuric acid evolves a perceptible vapour. 
* Proc. Roy. Soe., Nov. 16, 1876, No. 176, p. 304. 
f Do. Do. p. 305. 
+ Do. Do, p. 306, 
