268 VETTERSSON, ON WATER AND ICE. 



Lg — Qq = — 0.0004598 jjpg = — 0.0003776 y^j^^^^^ 



ti = — 4°\ ^ 



, __3°f"ClPl ^U.0000042 CC. 



and 



ti = — 4"! pi 



':} 



0.006698 CC. 



t, =— 1°\ 

 t = 0'| 



t = — 3") ti — t 



\v = 0.0001740 [Reg] .... w = 0.0001762 [Wiilln] 



Difference . . . . = I.2 % . 

 But it must be observed, that afterwards, when the voliimes 

 of the liquid are determined with the same instrument, the- 

 relation 



Lg — Qq = — 0.0004598 etc. . . . 

 liolds good no longer. The volume of the water is diminished 

 by the melting, and consequently the vohime of mercury in 

 the instrument is no longer Q but Qi. The above values for 

 liquid water then become 



Lg — Qiq = — 0.0010409 j^gg = — 0.0009054 svalin. 



qpi = 0.000000 



Pl— P r. 



T r = — 0.0017132 



Il — h 



w = — 0.0000767 [Reg] .... w = — 0.0000748 [Wulln] 



Difference = 2.5 %. 



If the dilatometer is thus adjusted, so that for frozen water 



Lg — Qq = O (or nearly 0), 

 it will perform its functions under more unfavorable ' circum- 

 stances, if afterwards applied to determine the expansion of 

 the liquid water formed by melting of the ice. 



The change of volume caused by the melting process is- 

 naturally independent of all irregularities of this kind, be- 

 cause the instrument is brought back again to its original 

 temperature after the melting is finished. The operation of 

 melting may therefore be considered to be accomplished at 

 constant temperature. 



The principle of the artifice just described was originally 

 invented by Pliicker and Geissler. 



C. Determination of the latent heat of ^water. 



The method invented for this purpose is minutely de- 

 scribed in Journ. f. prakt. chem. [2] Bd. 24 p. 151 and must 



1 I have chosen here tlie most unfavorable example, -which can occur,. 

 viz. the determination of the dilatation between 0^ and — 1°. 



