A MM ON I A- COB ALT BASES. 65 



each uniting with a single equivalent of oxygen to form a uniacid base. The 

 occurrence of a deutoxide of chlorine as a couplet is not more remarkable or more 

 improbable than that of deutoxide of nitrogen ; but experimental evidence is still 

 wanting to support the view which we have taken of the constitution of Raewsky's 

 bases which have as yet been very imperfectly examined. Claus 1 has also applied 

 the theory of conjugates to several of the ammonia-platinum bases, but has not 

 considered the subject from precisely the point of view which we have taken, 

 though his ideas, in the main, are the same. It is in the explanation of the diffex*- 

 ence in the saturating capacity of the various bases that Claus' view appears to us 

 less satisfactory than that which we have proposed. It is scarcely necessary to 

 remark that our theory applies to all bases containing ammonia and a metallic oxide. 

 We may, however, observe that it also harmonizes perfectly with the ammonium 

 theory if we consider ammonium as a conjugate hydrogen, or as represented by 

 the formula NEQs. 



We will here remark that our view of the theoretical constitution of the 

 ammonia-cobalt bases was distinctly enunciated in the paper already referred to, 

 as published by one of ourselves, in 1851. 



A glance at the formulas of the ammonia-cobalt bases, suggests the possibility of 

 generalizing the results which we have obtained, by two distinct methods. In the 

 first place, it is evidently possible, theoretically at least, to replace one or more 

 equivalents of ammonia in these compounds by an equal number of equivalents of 

 a compound ammonia, as for example, by methylamin, ethylamin, &c. Thus there 

 may be, for example, a species of Eoseocobalt, having the formula 



5N(C 2 H 3 )7Co 2 , 



which would represent the base which we have described under that name, in 

 which methylamin replaces ammonia. We have not yet been able to investigate 

 this point with any degree of thoroughness, and will here only mention, that an 

 experiment made by adding a solution of piperidin to one of chloride of cobalt 

 and allowing the solution to stand for some time in a half-filled flask, with frequent 

 agitation, led to no decisive result. We selected piperidin for this experiment, 

 because this alkaloid is comparatively easy to prepare, and does not itself oxidize 

 by exposure to the air. 



In the next place, the results of our investigation may be generalized by replacing 

 cobalt by other metals. This, as we have already remarked, has been done by 

 Claus, who obtained ammonia-rhodium and ammonia-iridium bases corresponding 

 to Roseocobalt, and like this, triacid bases. To judge, however, from the imperfect 

 notices of Claus' papers which have hitherto reached us, his compounds are not 

 isomorphous with those of Eoseocobalt. 



We have ourselves made many experiments in this direction, though as yet 

 without interesting results. Iron and manganese promised to afford similar classes 

 of compounds, yet in their behavior toward ammonia and oxygen the proto-salts 



Chemiseh-Pharmaceutisches Central Blatt No. 25. p. 189, Oct. 1854. 

 9 



