[satterly] 



CROOKES RADIOMETER 



79 



Fig. 3 curve AB is a sample of the curves obtained with the 

 larger iron ball. 



/6 20 24 £S 3^ 36 



/V= fiPM or CROCKCS' RADiOMZTLR 



<0 



'^PT 



The distances were measured from a vertical plane about \ 

 radius of the ball back from the front point to a vertical plane about | 

 length of the arms of the mill in front of the axis of rotation. 



For small distances the points lie on a straight line (herein 

 differing from Crookes' results). For large distances the points 

 diverge from the straight line (herein also differing from Crookes) ; 

 they lie on a curve which turns towards the intensity axis. The 

 R.P.M. decrease in greater ratio than the intensity and when the 

 intensity reaches a certain low value the vanes do not rotate at all. 



The equation of the curve obtained in one case {AB, Fig. 3) is 



— =.000025 .¥+.000075X10--^^^ 



There does not seem to be any special meaning to the exponential 

 term and the explanation of the departure from the straight line has 

 still to be made. It cannot be due to any departure of the calculated 

 intensities from the real values for at these distances (80 cms. and 

 over) the inverse square law could be applied to an iron ball of the 

 size used. 



To check this point an Ediswan pointolite source was used. 

 This gives the graph CD of Fig. 3, which is also a straight line 

 except near C. 



