el 
PTO terete Se Ot 
EAGER ERD GEOL ALS TE 
OF SEA-WATER.. 229 
facts only shew what decompositions may. occur from evapora- 
tion or other processes ; they do not prove-what the actual 
state of combination is in the original solution. 
It is obvious, that: this is merely a case belonging to the 
more general question, What is the state of combination in a 
compound saline-solution, and on what principle can it be de- 
termined what are the binary compounds that really exist in 
it ?—a question of considerable importance, but one at the 
same time of very difficult determination. 
- When it is admitted that this. cannot be inferred with cer- 
tainty from the actual products of analysis, the next most ob- 
vious view is, that it may be inferred from a knowledge of the 
real forces of affinity, as, according to these, certain binary 
compounds must be formed; and as the state of the science 
does not afford any certain estimate of the strength of attrac-. 
tion, the problem, it may be concluded, is at present incapable 
of being solved. 
This conclusion, however, is: by no means certain. Attrac- 
tion is so much modified in its operation by external forces, 
and combinations are so frequently established from the in-. 
fluence of these, that it is not clear that we should be able to 
determine what combinations would exist in cases similar to. 
those connected with the present investigation, from a know- 
ledge of the degrees of attraction, were we even in possession 
of it. It is rather, perhaps, from a knowledge of the influence 
_ of these external forces, that an approximation to the solution 
of the problem is to be attained ; and an extension of the prin- 
ciple I have illustrated in the preceding part of this paper, it 
appears to me may throw some light on the question. 
If the foree of cohesion has so much power in modifying 
chemical attraction, as to change its results, and establish com- 
_binations independent of the relative degrees of strength with 
which 
