102 



senting a remarkable difference of solubility, and other differences 

 not less striking. 



To remove every doubt, the bromide, obtained by treatment of the 

 brominetted bromide with triethylphosphine, was converted into the 

 corresponding iodide, which in its properties and composition was 

 found to be identical in every respect with the characteristic iodide, 

 which I have fully described in my last note upon this subject. 



The transformation of the monatomic into what I have represented 

 as the diatomic compound being satisfactorily established, the con- 

 clusive experimental demonstration of the diatomic nature of the 

 latter presented itself without difficulty in the conception of bromides 

 containing at once phosphorus and nitrogen, the molecular expression 

 of which would no longer admit of division. 



This class of dibromides actually exists ; they are readily produced 

 by submitting the bromide of the brominetted body to the action of 

 ammonia or monamines instead of triethylphosphine. 



I have formed as yet only three representatives of this new class of 

 bodies, which I propose to designate as phosphammonium-com- 

 pounds ; their examination is sufficient to fix the character of the 

 class ; it would have been easy to construct scores of similar bodies. 



Action of Ammonia upon the bromide of the brominetted body. 



The two substances, especially when in alcoholic solution, unite 

 with evolution of heat 



[(C 1 H 5 ) s (C 1 H 4 Br)P]Br+H J N= 



(C.H.). 



(C,H 4 )'' 

 H, 



N 



Br, 



Both the bromide and the corresponding chloride are very soluble, 

 and little adapted for analysis ; I have therefore fixed the nature of 

 this body by the preparation and analysis of the platinum-compound. 

 For this purpose the bromide generated in the above reaction was 

 treated with oxide of silver ; it is thus converted into a powerfully 

 alkaline solution obviously of the dioxide, 



(C 4 H 5 ) 3 

 (C 4 H 4 )" 

 H, 



P 



N 

 H 2 ~ 



