Analogies observed in Chemical Unions. 367 



compound is rather a hydrobase than a hydracid ; uniting with 

 oxygen acids by the intervention of water, and with anhydrous 

 hydracids. As it combines, however, also with ammonia, 

 there is, in some measure, a double type. The approach to 

 inorganic types is manifest in all the three series, although, 

 as usual, in a variable degi*ee of perfection. 



The singular combination kakodyle C* H^ A**, is not formed 

 on the type of the halogens, but rather of some of the ordi- 

 nary simple metals. We have its combinations with oxygen 

 and with simple halogens, quite on such a type ; and we have 

 itself in a separate state, although not at all comparable to a 

 metal in physical qualities, yet closely allied in chemical re- 

 lations. 



We have perhaps examples of imperfectly developed series 

 in the combinations of formyle C' H, and of acetyle C* H^. 

 But the evidence as to these, is as yet much less satisfactory 

 than in regard to Benzoyle. 



The ordinary organic acids appear to belong essentially to 

 the type of oxygen acids, as the organic bases belong to the 

 type of hydrobases. They unite directly with oxybases, and, 

 in combining with hydrobases, the latter change their type, 

 through the intervention of the elements of water, as we see 

 in the compounds of organic acids with ammonia and with 

 ether. But neither in the case of the acids, nor of the bases of 

 organic chemistry, does it seem to be essential that we should 

 hold they actually contain compound radicles as such. The union 

 with oxygen in the one case, and with hydrogen in the other, 

 may be determined by a principle of analogy ; and the chemi- 

 cal character of the resulting compound may be determined by 

 the analogy ; but the actual grouping of elements may vary 

 from the type. We are prepared, by the principles which I 

 have been illustrating, for such imperfections in the results. 

 I subjected to voltaic action fused benzoic acid and fused 

 morphia, without any symptoms of decomposition, and even 

 without the galvanometer being affected ; there thus being no 

 evidence of binary combination. 



hJome leading combinations of elementary bodies amongst 

 themselves, have thus been briefly followed out into tlie less 

 perfect analogies of compound or organic radicles. It is very 



