364 Professor Connell on the 



it appears to pass by an easy transition from inorganic to organic 

 nature ; and through, the whole series we observe a strong 

 analogy of relations. Next to ammonia, we may reckon cer- 

 tain hydro-carbons ; and, first, the combination C* H*. This 

 unites with the hydracids, and forms the haloid ethers, 

 such as muriatic ether, C* H*, CI H. To combine with oxy- 

 acids, it must, like ammonia, change its type ; and it accom- 

 plishes this in the same way, through the aid of the elements 

 of water. It thus becomes common ether, C* ff 0, and then 

 unites with oxyacids to form the ordinary saline ethers, such 

 as hyponitrous ether, C'* H^. + N 0^. All these combina- 

 tions, although formed on the type of the ammoniacal com- 

 pounds, are of an imperfect character as saline bodies, the 

 different acids not producing in the compounds their ordinary 

 reactions. Although the analogy has determined the union, 

 yet the result is of an imperfect character as a salt, the acid 

 and alkali appearing not to have ultimately united as such. 



Analogy would lead us to infer the existence of a compound 

 metallic substance, C^ H^, analogous to ammonium, but no other 

 evidence than inference from its combinations has yet been 

 brought to light. I tried to decompose ether by direct voltaic 

 agency, and likewise to form an amalgam, by galvanizing a 

 globule of mercury placed in alcohol, but without success. It 

 is probable, therefore, that although ether is formed on the type 

 of the metallic bases, yet the actual result is modified, so as 

 to have as little the constitution of a true oxide, as its combi- 

 nations with acids have of true salts. 



In all such unions, we may bear in mind, that although the 

 analogy or type may determine the union, as, for example, be- 

 tween the elements of ethyle and oxygen, and between ether 

 and an oxyacid, yet the compound actually resulting may not 

 contain the radicles as such ; but a new arrangement of elements 

 may take place, and the result be, to that extent, imperfect. 

 We may consider oxide of ammonium as forming the transition 

 between perfect oxides and those still less perfect than the 

 connecting link. 



The hydro-carbon C^ H^ affords the pyi'oxylic series, on en- 

 tirely analogous principles to the alcohol series. 



Another series of this kind, not yet fully developed, appears 



