Manchester Memoirs, Vol. xliii. ( 1 899), No. 4. 13 



theory, it may with justice be urged that the escape of 

 gases from the earth's surface is slow, though of course 

 more rapid in the case of the light than of the heavy gases ; 

 and that if there is a constant emission of hydrogen or of 

 helium from the earth, it may be replaced more quickly 

 than it escapes. And the fact that the decomposition of 

 animal and vegetable matter is sometimes accompanied 

 by the evolution of hydrogen, and that many hot springs 

 evolve gas comparatively rich in helium, may perhaps 

 account for their occurrence in the atmosphere, in spite of 

 their low densities. 



Be this as it may. Dr. Stoney's theory was a consider- 

 able factor in inducing Dr. Travers and myself to again 

 examine argon, with a view of finally settling the question 

 of its homogeneity 



Our patience, however, was not equal to the task of 

 fractionally diffusing argon. The experiments on helium 

 lasted over three months, and they involved an immense 

 expenditure of mechanical labour, unenlivened by any 

 results. Moreover, the rate of diffusion of argon is only 

 half as rapid as that of helium, and a similar set of ex- 

 periments would have taken more than twice as long a 

 time. But the most cogent reason against employing the 

 process of diffusion to effect a separation was, that although 

 by its help, it is possible to separate two gases of widely 

 differing density, if each is present in considerable quantity, 

 it is not an advantageous process, if it be required to 

 separate two gases not differing widely in diffusion rate, 

 and of which one must be present only in minute amount. 

 It will be remembered too that the high density of argon 

 rendered it much more probable that it would contain a 

 small quantity of a heavy gas, than of a light one. 



Previous experiments had made it certain that crude 

 argon would contain only a small amount of admixed 



