40 



Mr. William Croohes 



[Feb. 18, 



process, to be effected only in a very high vacunm. To understand 

 this process it is necessary to make an apparent total digression. 



It seems, perhaps, strange to speak of exhausting the air in hollow 

 bnlbs and tubes until there is left in them only the one-millionth 



Fig. 1. 



FRACTIONATION 



OF YTTBIA. 



J) ca) (2) « e.. 



@ @ @ @ ® ® @ @ @ 

 ^®x®x®x®x®x®x®x®x®x®x®x®x^x-x-x 



^ ®@® 00 @®® 00 01 00-0 



^^©^@V®V@^@^®V-0V®^5VVV@ 



II illil II IMI! !I Hill II SMII 1 



SPECTRA OF FIVE COMPONENTS OF YTTRIA 



part of an atmosphere. It is only in modern times that atmospheric 

 air has come to be regarded as matter. To this day a bottle or a jar 

 is said to be " empty " if it contains no liquid or solid body, the air 

 with which it is filled being completely ignored. According to the 

 same common idea, how empty then must a vessel be when the air it 

 contains is reduced to the one-millionth part of its original quantity ! 

 That something still remains is, however, proved by the fact that I 

 have succeeded in reducing the pressure down to one fifty-millionth 

 of an atmosphere. What this number represents will be better under- 

 stood if I say that, given a barometric column one hundred miles in 

 height, the remaining pressure would be equal only to about the tenth 

 of an inch. Even this high degree of exhaustion by no means 

 represents an absolute vacuum. I have in this glass tube perhaps the 

 nearest approach to perfect emptiness yet artificially obtained. Its 

 internal capacity is 5 cubic centimetres, and it is exhausted to 

 the one fifty-millionth part of an atmosj)here. It still contains 

 100,000000,000000 molecules. The internal space, therefore, is far, 

 very far, from being absolutely void of matter. 



