1346 



into Vso degree) in the experiiuents of table ^. Consequently the 

 agreement of" the values in the last column but one oi' table 6 is 

 much better: the mean values of the three tables are in good 

 concordance. 



When we now determine the mean of the \alues from the last 

 columns of tables 4 — 6, the tirst value of table 5, which is evidently 

 too high (probably because the transformation in the quadruple point 

 had not yet entirely taken place), being eliminated, the value 14270 

 for E^ is found; when the value in question is taken into account, 

 the mean amounts to J4350 cal. 



Hence the decomposition of the hydrate takes [)lace according to : 

 II^S . n H,0-^h^S -f- u /y,0 — 14270 (14350) ral. 

 (solid) (<7"*) {liquid) 



An objection that may be adduced against the above calculations, 

 is the choice of ii = (3 in the determinatixJu of r and .9 from equa- 

 tion 10 (I and h, and in the calculation of the algebraic sum of 

 the specific heats. We shall come back to this in § 11- 



10. From the values E, =5550 cal. (^ 8) and E, = 14270(14350) 



cal. (§ 9) follows according to equation 1 : n Q = S720 (8800) cal., 



and 'as Q=J440 cal., the value 6.06 (6.11) follows for n. Hence 



the conclusion from these calculations is, that the hydi-ate has the 



formula: 



H,S 6H,0. 



11. As was stated in § 9, /i :^ 6 was alreaily taken in the 

 calculation. A choice of n. was necessary to render it possible to 

 calcidate r and s (equation 10 <i and h), and find the algebraic sum 

 of the specific heats. We shall, therefore, still have to show that 

 ?i =z Q is the oidy value that satisfies the observations. It might, 

 namely, also be possible that with the' choice ;/ = 5 the result of 

 ^ 10 was also appreciably changed, and would correspond to the 

 choice 11 = 5. This is, however, not the case. The following consi- 

 derations may make this clear. 



If n =z 5, the molecular specific heat of the hydrate would amount 

 to 53.0, and as the specific heat of five molecules of water and one 

 mol. of sulphuretted hydrogen is 90 -)- 6.3 = 96.3, the algebraic 

 sum would be about 43. 



When on assumption n = 5 we calculate the \'alues of /• and .s-, 

 and by the aid of this the other values, it appears that this does 

 not give an} change in the result, and that, therefore, the assump- 

 tion ?i =: 5 IS erroneous. We have carried out this calculation for 

 the data of table 4; the results are recorded in table 4a. The mean 



