186 



Mr. W. W. J. Nicol on the Molecular 



water of crystallization or are anhydrous ; the second those 

 which contain water of constitution as well as water of crys- 

 tallization. It is to be noted that those in the former part 

 of the table agree pretty closely, even though the quantities 

 of water of crystallization are so very d liferent, and that those 

 in the latter part also agree among themselves, but the value 

 of A is about 3 units lower. 



Table V. 



No. in 

 Table II 



M. 



n. 



w(MCl). 



w(M^- 4 ). 



Diff. 



A-—- 



— n 



Aq. 



1- 5. 



K 



1 



1827-7 



1818-8 



8-9 



8-9 



0-0 



15-19. 



Na 



1 



1817-5 



18083 



9-2 



92 



0-10 



39-46. 



Cd 



2 



2 



36278 



3610-1 



17-7 



8-9 



2-1 



43-47. 



Li 



1 



1817-9 



1808-1 



9-8 



9-8 



1-1 















9-2 





42-50. 



Cu 



2 



2 



Magnesian 



Sulphates. 



122 



61 



2-5 



36136 



3601-4 



41-49. 



Co 



2 



2 



3613-4 



3600-6 



12-8 



6-4 



6-7 



40-48. 



M 

 2 



2 



3611-1 



3595-5 



15-6 



7-8 

 6-8 



6-7 



i 



Now we have seen that water of crystallization does not 

 affect the molecular volume to an extent that can be recog- 

 nized by experiment ; the cause of this difference in the case 

 of the magnesian sulphates must therefore lie, not in the 6 or 

 4 molecules of water of crystallization, but in the single 

 molecule of water of constitution. From the above experi- 

 ments it therefore appears probable that water of constitution 

 can be recognized when the salt is dissolved ; I say probable, 

 for owing to the difficulty experienced in preparing solutions 

 of these hydrated salts of the exact composition required, 

 some of them parting with their last molecule of water at a 

 temperature little below that at which the salt begins to 

 decompose, the results are not so concordant as in the case of 

 the other salts. 



There is, however, another method by which this point can 

 be examined, and one which is totally independent of the 

 molecular composition of the solutions employed ; it is that 

 of determining the volume-change on double decomposition. 



