﻿304 Mr. Baly and Dr. Ramsay on the Relations of Pressure, 



considers to be rendered probable by Regnault's own work. 

 For he, too, found that air is up to a certain degree more 

 compressible at pressures higher than that of the atmosphere. 

 For example, if Regnault's results for pressures between 3 

 and 6, between 6 and 12, and between 12 and 24 atmospheres 

 be termed k\ k", and k' /f respectively, then 



jfe"/# = 0-997128 ; k'"/k" = 0*999386. 



The question of course is, Can these and similar ratios 

 increase to above unity ? 



Siljestrom found from similar but less numerous experi- 

 ments with oxygen and with carbon dioxide that the variation 

 in the ratio (V + Y')/V was in the same direction. His results 

 with hydrogen are :— 1-47247 at 759-351 millim. ; 1-47302 

 at 351-162 millim.; 1*47258 at 162-75 millim.; and 1-46895 

 at 75-17 millim. 



Siljestrom's results are criticised by Mendeleeff and by 

 Amagat. Any error which he might have made is cumulative, 

 as pointed out by Van der Yen, whose work will be sub- 

 sequently considered. 



MendeleefPs work is at present unfinished, or at least a 

 description of it in full has not yet appeared. His main work 

 appeared in Russian in 1875, and the first part consists of 

 267 pages folio ; the second part is not yet published. We 

 owe to the kindness of Dr. Jas. Walker an account of the 

 contents of this large volume. Short accounts of MendeleefFs 

 general conclusions are to be found in French in the Melanges 

 Physiques et Chimiques tires du Bulletin de V Academie Im- 

 perial des Sciences de St. Petersbourg, ix. ; also Annates de 

 Chimie et Physique, [3] ix. 1876. These abstracts leave much 

 to be desired. 



Mendeleeff slates that he tried nearly a dozen different 

 forms of apparatus before a satisfactory one was devised. 

 This, he says, is figured in volume ii. table 15 (unpublished)* ; 

 but he gives the following description of it in a lecture 

 delivered before the Imperial Russian Technical Society in 

 January 1881. It consisted of an ovoid glass vessel, with a 

 capacity of about 3^ litres, placed in a trough of water. The 

 lower portion of this vessel terminated in a tube, one metre in 

 length, provided at its lower end with a stopcock ; the volume 

 of air was determined by weighing the mercury run out from 

 this vessel. A tube was sealed to this vertical tube, which 

 served to fill the measuring-vessel with mercury when re- 

 quired. A capillary tube was attached to the upper part of th< 



* Professor Mendeleeff informs us that a large part of the secon( 

 volume is now in manuscript. 



