108 Prof. D. Mendeleeff on the Variation in the 



and myself, and showed that beyond 100° the expansion of 

 water increases just the same as at lower temperatures. Later 

 on Hirn * accurately determined the variation of a volume of 

 water heated from 100° to 200°, allowing it to remain all the 

 time under a pressure of about 15 atmospheres. Taking the 

 volume at 4° to be unity and taking, according to Despretz, 

 V, = 1-04315 for 100°, it appeared that, for 



120° 140° 160° 180° 200° 



V*=l-05i*92 1-07949 1-10149 1-12678 1-15777 



In order to render these figures comparable with the other 

 data for a pressure of one atmosphere, it is necessary to 

 multiply them by (1 4-/^14), since the volumes were observed 

 under a pressure of 15 atmospheres f. This reduction neces- 

 sitates the knowledge of fi t between 100° and 200° inclusive. 

 Up to the present time direct determinations of this kind do 

 not exist, so that it becomes necessary to extrapolate by means 

 of formula No. 7. This gives for the above temperatures : — 



10 6 /^=46-17, 52-73, 61-37, 73-09, 84'89. 



Therefore Hirn's figures for the volumes of water when reduced 

 to a pressure of one atmosphere become : — 



Y t = 1-06060, 1-08029, 1-10244, 1-12793, 1-15914. 



Determining from these the density under a pressure of one 

 atmosphere, we have 



S,=0-94286, 0-92568, 0*90708, 0-88658, 0-86271. 



like Sorby's, were only intended to give a preliminary acquaintance with 

 the phenomenon, and my error is still greater than Sorby's, namely about 

 HhO'01. For water a determination was made for three temperatures, 

 and gave the following results : — 



£=120° 140° 160°. 



V=l-07 1'09 111. 



The volumes were reduced to a pressure of 1 atm. Sorby's and my 

 results are incomparably less accurate than those made by Hirn, aud as 

 such have not met with any further attention. This was a first recon- 

 naissance into the region of the unknown. 



* Hirn, 1867, Ann. de Chimie et Phys. (4) x. p. 32. The method of 

 determination and the dimensions of the vessels adopted guarantee con- 

 siderable accuracy to Hirn's results, which, however, judging from the 

 mode of computing the corrections, especially for the coefficient of 

 expansion of the vessel, contain an error hardly less than +O0005. 



t Hirn states in his memoir (7. e.) on p. 39, that the height of the 

 mercury in the open column was 11-25 metres ; but on p. 48 he says that 

 the mean pressure was equal to 11*5 metres. Taking the first statement 

 and adding the atmospheric pressure, we obtain 15*8 atm., but if we take 

 the height 11*5 metres to express the total pressure we obtain 151 

 atm. 



