Analogies observed in Chemical Unions. 363 



reasons, it is less perfect than potash or soda, considering it 

 as belonging to the series of oxyalkalies. 



In this way we have both the alkali ammonia and the alkali 

 oxide of ammonium ; the leading tendency of the one being 

 to combine with hydi'ogen acids, and of the other with oxygen 

 acids ; but the one being, in all its relations, of a more perfect 

 character than the other. They belong to entirely different 

 types, the one being the most perfect of hydrogen bases, and 

 the other being one of the less perfect of oxygen bases, and 

 having a compound radicle. 



' Its radicle ammonium may be viewed as an imperfect mem- 

 ber of the metallic series. It is formed, in respect of properties, 

 on the type of the metals, but has scarcely a substantive ex- 

 istence. Facts, however, appear to tell us, that it is not alto- 

 gether imaginary. The metallic apjDearance of the ammoniacal 

 amalgam leads strongly to the idea, that the mercury has com- 

 bined with something which does not mineralize it, but is 

 metallic like itself. And when this occurrence is taken in con- 

 nection with the usual constitution of the oxysalts of ammonia, 

 and their bearing on isomorphism, we certainly see strong 

 grounds for holding ammonium and its oxide to have an ex- 

 istence, although of a more or less temporary nature ; and, 

 therefore, imperfect, as regards other metals and bases. 



It appears to me, that it ought to be considered as enter- 

 ing into our view of the natural tendency of hydrobases to 

 combine with hydracids, that the primary affinity may often 

 result in such an arrangement of the constituent atoms, that 

 a haloid salt of ammonium instead of a hydracid salt of ammonia 

 shall be formed. There are evident traces in these compounds 

 of a double type, and of a ready transition of the one into the 

 other ; and it is not always easy to say when the one arrange- 

 ment and when the other actually takes place. It is really 

 not of great importance whether the atom of hydrogen is in 

 more immediate connection with the halogen or with the other 

 elements of the compound metal. It may oscillate, so to speak, 

 between the two, and take its ultimate position according to 

 circumstances, as it appears to have affinities for the latter as 

 well as the former. Nothing is thus added nor taken away. 



The series of hydrogen bases is a pretty extensive one, and 



