INFINITESIMAL QUANTITIES RAMSAY. 227 



First, the tube was of glass, and its density diffei-s from that of 

 silica of which the counterpoise consisted. 



To determine the alteration due to this difference of densities, 

 the bulb was removed and a counterpoise of solid silica substituted. 

 Weighuig agam, we fomid a difference of 91 millionths, which being 

 subtracted leaves 561 millionths. Secondly, a correction is to be 

 made owmg to the xenon's having been weighed at a pressure of 

 70 milhmeters, while the j^ressure has been changed to 52.9 milU- 

 meters. If we had weighed the sealed tube at a pressure of 52.9 

 millimetei-s instead of 70 milhmeters, it would have weighed more. 

 The correction, then, is positive. We must find the difference 

 between the Avcight of 0.536 cubic milhmeter (the capacity of the 

 bulb) at 70 and at 52.9 miiUmeters-(17.1 milhmeters). The equation 



17.1 X0.536x4^X 1000= 15 milhonths 

 7o0 



gives this weight, where 1.29 is the weight in milligrams of a cubic 

 centimeter of air. 



Thus we have 561 + 15 = 576 millionths of a niilligram for the 

 weight of the xenon. It is easy to show that the weight calculated 

 should be 577 milhonths. 



Allow me to give another example of weighing with this balance. 

 Mr. '\Yhytlaw-Gray and I have ascertained the weight of hehum 

 formed by niton vv'hen it changes to radiimi A, B, C, and lastly D, 

 In the mamier already described we filled a tube with niton 

 in July, 1910. We left it mitil the beginning of October, that it 

 might change to heUum and radium D. This last has a half -life 

 of about 16 years, so that it may be regarded as permanent. The 

 tube was weighed, the point was broken, and it was immediately 

 again placed on the balance. The loss of weight was 15 milhonths, 

 its volume was 0.196 cubic milhmeter, and the weight of air which 

 entered at a pressure of 37.7 milhmeters and 18.5° C. was 12 mil- 

 lionths; therefore, the total weight of the helium was 27 milhonths. 



But given the cjuantity of niton employed, v/e should have obtained 

 38 milhonths, so it became necessary to look for it. Now, under the 

 influence of niton the gas molecules which are found in the same 

 vessel are forced to accelerate their velocity, doubtless because of 

 collisions with the a particles projected during the disintegration of 

 the atoms of niton. This velocity causes them to penetrate the walls 

 of the vessel containing them, and the result of our experiments is 

 that the penetration depends not only upon the velocity but also 

 upon the size and form of the molecules. Thus, helium mixed with 

 niton penetrates the glass more than neon and neon more than 

 hydrogen. However that may be, we heated the tube in a small tube 

 of silica surrounded by a second tube to prevent the entrance of the 



