101 



It is true C 2 H 2 figures in this formula as monatomic, whilst we 

 should expect it endowed with diatomic substitution-power. But 

 the connexion between composition and substitution-power is by no 

 means finally settled ; in fact, we know of many cases in which, 

 under conditions not sufficiently established, the atomicity of a mole- 

 cule changes : witness the radical " allyl," which is capable of re- 

 placing one or three equivalents of hydrogen. 



But without going this length, the scission of diatomic ethylene 

 into two monatomic molecules may take place in many other ways. 

 The transformation of dibromide of ethylene into hydrobromic acid 

 and bromide of vinyl, 



(C 4 HJ" Br 2 =HBr + (C 4 H 3 ) Br, 



is a familiar example. The splitting of the ethylene-compound into 

 bromide of formyl and bromide of methyl, 



(C 4 HJ" Br 2 = (C, H) Br+ (0, HJ Br, 

 has never been observed, but did not appear altogether unlikely. 



Our analytical methods are insufficient to distinguish between 



[(C 4 H 5 ) s (C, HJ P] Br and [(C 4 H 5 ) 3 (C, H) P]Br ; 



and what I have represented as a diatomic ethylene-compound 

 might have been, after all, a monatomic bromide the bromide of 

 formyl-triethylphosphonium, the complementary methyl-compound 



[(C 4 H 5 ) 8 (C 2 H 3 )P]Br 



existing possibly among the secondary products of decomposition. 



In the presence of these and several similar self-raised objections, 

 bv which every observer endeavours to test the truth of his con- 

 clusions, I was induced again to appeal to experiment. 



The prosecution of this line of the inquiry has led me to the disco- 

 very of a new class of diatomic bodies, which, while it confirms incon- 

 testably the correctness of my interpretation, appears to claim the 

 attention of chemists for several other reasons. 



I have established, in the first place, that the monatomic bromide 



[(C 4 H 5 ) 3 (C 4 H 4 Br)P]Br 



may be readily converted into the diatomic bromide 

 [(C 4 H 5 ) 6 (C 4 H 4 )''PJBr 2 



by the simple addition of triethylphosphine. Nothing is easier than 

 to prove the transformation, the platinum-salt of the two bases pre- 

 VOL. x. i 



