418 Prof. A. P. Chattock on the Velocity and 



wire (diam. 4*7 cm., thickness 0*07 cm.) and placed the point 

 at its centre (fig. 4). If under these conditions the electricity 

 were to arrive at the ring equally on both sides, the pressure 

 measured on the gauge would be it. It would, however, 

 certainly travel most to the side remote from the point, as 

 indicated by the dotted lines, and the pressure would thus be 

 too large ; but by repeating the measurements with a point of, 

 say, larger diameter, at which the discbarging-field would 

 therefore be less, the difference between the pressure-readings 

 should aive the change in it due to the change in field. 



Points of diameters O0066 cm. and O0306 cm. respectively 

 were used for this purpose ; their respective discharge-fields 

 for positive electricity being 1600 and 640, and for negative 

 1280 and 490 E.S. units. It was necessary to take a large 

 number of readings as the pressure-variations mentioned above 

 were of course again met with, and, the pressures themselves 

 being extremely small, these variations were rather serious. 

 Thirty- five observations with an average current of 7930 E.s. 

 units gave mean pressures of 3"9 dynes with the thin point 

 and 3'3 with the thick for positive discharge ; the corresponding 

 pressures for negative discharge being 2*65 and 1'50 re- 

 spectively. Substituting in the above equation for e the 

 values 3-9-3-3 for tt, 1600-640 for F, and 7930 for C, we 

 obtain 22 as the number of molecules in the diameter of a 

 positive cluster. For the negative cluster the number is 27, 

 using the values of it and F for negative discharge. 22 agrees 

 well with the previous 19; a good deal better, in fact, than the 

 many sources of error would lead one to expect ; but the 

 number for the negative clusters is likely to be too large, as the 

 negative discharge from the blunt point formed a luminous 

 projection which stood out from it farther than that from the 

 sharp point. If this represented a line of dissociated gas it 

 would behave like a conductor, and the pressure-producing 

 part of the discharge would thus be starting from the outer 

 end of it. This would make the pressure too low ; and the 

 change in pressure, on changing from blunt to sharp point, 

 too high. 



It is a pity that there is no check to be had to this from 

 the negative curve in curves III., which appears to give a 

 negative and therefore impossible value to it. The probable 

 reason for this is, however, an interesting one on its own 

 account. About two years ago, in measuring the field in air 

 at a sharp discharging-point, I found that, in the case of 

 positive discharge, the field was almost constant for the widest 

 range of current I was able to apply when discharging onto a 

 smooth cathode of large area ; but that if even a small amount 



