﻿Volume, and Temperature of Rarefied Gases. 307 



and becomes more compressible. This is a very strange 

 conclusion, but it Las been confirmed by our work. 



In criticising Bohr's work, it is to be noted that he em- 

 ployed very wide tubes, one as barometer, the other to contain 

 the rarefied gas. His method of reading was ingenious. 

 Finding it impossible to read the level of the mercury in such 

 wide tubes with accuracy, he placed in each a glass ball, and 

 read its upper surface. The readings therefore appear to be 

 very accurate. But the shape of the meniscus of mercury is 

 different in a nearly perfect vacuum and in a vessel contain- 

 ing gas ; it is always flatter in the vacuous tube. Hence the 

 volume read was probably always too small ; and increasingly 

 smaller as the gas was more compressed. This would render 

 the value of the product pv increasingly smaller, and necessi- 

 tate the correction he applies. It would, however, not explain 

 the breach of continuity which he observed. 



Next in order of time comes a paper by F. Fuchs (Wied. 

 Ann. xxxv. p. 430) in 1888. The lowest pressure which he 

 measured was 248 millim. ; but though he found deviations in 

 the case of sulphur dioxide, hydrogen was normal. 



In 1889, E. van der Yen carried out experiments by a 

 method nearly identical with Siljestrom's, but avoiding the 

 possibility of cumulative errors which Siljestrom's method 

 involved. His conclusion is opposite to that of Mendeleeff 

 and Siljestrom; it is i — If a volume of air under low pressure, 

 contained in a closed space, be doubled, it behaves as if its 

 pv decreased with decrease of pressure. Thus he finds 



At 62 millim. pressure, ^ = 0*9873, 

 At 31 „ „ ^=0-9811, 



At 16 „ „ ^ = 0-9740. 



Van der Yen used iron cylinders with walls 6 millim. thick, 

 and it is unlikely that any appreciable alteration in their 

 volume would be caused by the differences in pressure which 

 occur in his experiments. He appears also to have been 

 careful about the temperature of the vessels. But the latter 

 part of his paper is occupied with a discussion of the 

 capillarity of his mercury-gauge ; and it appears to us 

 that a possible source of error lies here. For the height of 

 the meniscus of mercury in a gauge is undoubtedly altered 

 by the pressure of the gas in contact with it; and as the gas 

 was rarefied, the apparent difference in height of the mercury 

 in a narrow r gauge like the one he employed would be in- 

 fluenced not merely by capillarity, but also by the form of 

 the meniscus. We regard it ns probable that the capillarity 

 of mercury, like that of other liquids, is altered by the gas in 



