62 RESEARCHES ON THE 



while the secondary radicals are 



5NlCCo 2 

 NOT5NH3?Co 2 0. 



The oxides on this view are 



5NlCCo 3 0.0 2 

 NO?5NlCCo,Q.0 2 



and are consequently of the form RQ 2 , so that their biacid character is explained. 

 The doctrine of polyacid bases is by no means new ; it is in fact contained in the 

 empirical law above referred to, that there are in neutral salts as many equivalents 

 of acid as there are of oxygen in the base, bearing in mind, however, that the 

 oxygen in the base must be outside of the radical. The ammonia-cobalt bases 

 like the conjugate metals produced by the union of ethyl, methyl, &c, with anti- 

 mony, arsenic, and bismuth, serve, however, to place the doctrine of polyacid bases 

 upon the same footing as that of the polybasic acids, so that the two theories are 

 in this way complementary to each other. From this point of view it is interest- 

 ing to remark, that the chlorplatinates and double cyanides of the ammonia-cobalt 

 bases follow the same law as the oxygen salts, thus we have 



6NH 3 .Co 2 Cl 3 +3PtCl 2 

 5NH 3 .Co 2 Cl.Cl 2 +2PtCl 3 

 N0 2 .'5NH 3 .Co 2 O.Cl 2 +2PtCl 2 

 6NH 3 .Co 2 Cy 3 +Co 2 Cy 3 



N0 3 .5NH 3 .Co 2 O.Cy 2 + FeCy. 



In point of fact, the presence of but two equivalents of bichloride of platinum 

 in the chlorplatinate of Purpureocobalt first led us to suspect that the true oxygen 

 salts of this base would be found to contain but two equivalents of acid. 



Two other points require special notice in this connection. We have already 

 shown that the oxide of Purpureocobalt, in at least two cases, is capable of uniting 

 with four equivalents of acid so as to form feebly acid salts. We consider these 

 salts the true bi-salts of the base, and not as double salts of Purpureocobalt and 

 water. In other words, we hold that they bear the same relation to the neutral 

 salts of the base which bichromate of potash does to the neutral chromate. If 

 this view be correct, we may perhaps expect to find salts of Eoseocobalt or Luteo- 

 cobalt containing six equivalents of acid. The only acid salt of Luteocobalt 

 hitherto discovered is the carbonate, but this in reality, in our view, is a double 

 salt of Luteocobalt and water, and has the formula 



6NH 3 .Co 2 3 ,3C0 2 + HO,C0 2 . 



The fact, that both the acid and neutral oxalo-sulphate of Purpureocobalt con- 

 tain two distinct acids, is also a very instructive one, since it completes the analogy 

 between the polyacid bases and the polybasic acids. A polybasic acid, as, for 

 example, tartaric acid, may unite with two different bases at once, and we now 



