Prof. Graham's Note on the Constitution of Salts. 219 



portant. It is also interesting to consider the conditions that 

 are necessary in order that a root should be wholly real or 

 wholly imaginary; and to observe the curious manner in 

 which, when X and v are both = 0, the solution here given 

 identifies itself respectively, according to certain relations be- 

 tween X and /*, with the ordinary algebraical or trigonometri- 

 cal solution. 



XXIX. Note on the Constitution of Salts. By Professor 

 T. Graham *. 



T^HE author may perhaps be excused in drawing the at- 

 -*- tention of chemists to a distinction in saline combina- 

 tions, which is at present too often overlooked, and confusion 

 thereby occasioned. The orders of monobasic, bibasic, and 

 tribasic salts, of which the phosphates proved types, have 

 lately been greatly enlarged by the discoveries of Liebig and 

 Dumas respecting vegetable acids, and the distinctive charac- 

 ters of these orders are well understood. The best proof of 

 an acid's being bibasic or tribasic, is its combining at once 

 with two bases which are isomorphous, or belong to the same 

 natural family, as phosphoric acid does with soda and ammo- 

 nia in microcosmic salt, and tartaric acid with potash and soda 

 in Rochelle salt. Water and magnesia, water and barytes, 

 water and oxide of lead, are also constantly associated as bases * 

 in bibasic and tribasic salts, but never in true double salts, or 

 combinations of two or more salts with each other, with which 

 salts of the preceding orders are apt to be confounded. 



But it is too generally supposed, that a metallic oxide can- 

 not exist in a saline combination, except in the capacity of 

 base, although in most of those bodies which are at present 

 termed subsalts, the whole or a portion of the metallic oxide 

 is certainly not basic, but is attached to a really neutral salt 

 in a capacity similar to that of constitutional water, or water 

 of crystallization. Oxide of copper, oxide of lead, barytes, 

 and the other metallic oxides included in or related to the mag- 

 nesian family, appear to rival water (which is a member of the 

 same family), in the frequency with which they discharge this 

 function in the constitution of saline compounds, particularly 

 of those belonging to the organic kingdom. Thus the neutral 

 organic principle orcine combines with five atoms of oxide of 

 lead, according to Dumas, which replace five atoms of water 

 which orcine otherwise possesses. But it should be brought 

 prominently into view, that neither the water nor the oxide of 



• Read before the Chemical Section of the British Association, at the 

 late meeting at Newcastle, Aug. 22, 1838; and now communicated by the 

 Author. 



