Mr Graham on Water as a Constituent of Salts. 299 



combine with water as a base in one proportion only, so far as is 

 yet known. By these discoveries in regard to phosphoric acid 

 and its salts, the ordinary conceptions entertained of the consti- 

 tution of salts were completely deranged. The salts called bi- 

 phosphate of soda, phosphate of soda, and subphosphate of soda, 

 are all tribasic salts. The common idea of a supersalt is inap- 

 plicable to any of them. 



I have subsequently found water to exist in a different state 

 in certain salts, not possessed of a true basic function, being re- 

 placeable by a salt, and not by an alkaline base. To illustrate 

 this new function of water as a constituent of salts, is my princi- 

 pal object in the present communication. 



The tendency of phosphate of soda to unite with an addi- 

 tional dose of soda, and form a subsalt, I had traced to the exist- 

 ence of basic water in the former. The inquiry suggested itself, is 

 there any analogous provision in the constitution of such salts as 

 have a tendency to combine with other salts, and to form double 

 salts ? The salts which combine together most readily are the 

 sulphates, and to these I therefore turned. The result was, that 

 in that well-known class of sulphates, consisting of sulphates of 

 magnesia, zinc, iron, manganese, copper, nickel, and cobalt, all of 

 which crystallize with either five or seven atoms of water, one 

 atom proved to be much more strongly united to the salt than the 

 other four or six, which last generally may be expelled by a heat 

 under the boiling point of water, while the remaining atom uni- 

 formly requires a heat above 400° Fahrenheit for its expulsion, 

 and seems to be in a manner essential to the salt. The consti- 

 tution of crystallized sulphate of zinc, for instance, may be ex- 

 pressed thus : 



ZnSH+H 6 



We here divide the seven atoms of water into one atom, 

 which is essential to the constitution of the salt as we know it, 

 and six atoms which are not so ; and to this last quantity we 

 may restrict the application of the name " water of crystalliza- 



p p 2 



