34 PROCEEDINGS OF THE AMERICAN ACADEMY 



most probable that one of the two additional bonds is satisfied by sili- 

 con, the other by fluorine, and that the following graphic formula re- 

 presents its constitution. It should be remembered, however, that 

 this formula is only the most probable one, for, as already stated, we 

 have been able to bring no absolute proof of its correctness. 



H g = -pg--C 6 H 6 

 C 6 H* - JN — F 2 — Si F 2 — — Si — F 2 — JN — C 6 V 



If this formula is adopted, the formation of the silicotetrafluorides 

 can be explained in the following way. In the first place one mole- 

 cule of the fluoride of silicon acts upon one molecule of the base to 

 give the group 



C 6 H 6 -ll -F 



and, although this mode of union seems strange at first sight, it is not 

 without analogy, if we consider the close relationship of silicon and 

 carbon, as then it is similar to the formation of ammonium salts by the 

 action of methyliodide on a base, as shown by the following re- 

 actions : 



C 6 H 5 NH 2 + CH 3 I - C 6 H 5 - IN - I 



H - = AT - SiF 3 . 



iF.,F = C a H s - IN - F 



C G H 5 NH 2 + SiF 8 * .= ^ 5 

 Since the group 



C fi H s -lN -F 



■^"s 



is at once a fluoride and a substituted fluoride of silicon, an action 

 next takes place similar to the formation of a fluosilicate from a fluo- 

 ride and fluoride of silicon, thus : 



C 6 H; - N - F 1 3 = C 6 Hs - ]N-F=F- Si = Fs ; 



K — F = F— CI _ v 



4 = K — F = F — Oi ~~ 2 * 



2 KF + SiF 



In this reaction the formation is assumed of the bivalent radical F 2 n , 

 which has been proved to exist in hydrofluoric acid by Mallet's de- 

 termination* of the vapor-density leading to the formula H 2 F 2 , and 

 the assumption of the presence of which in fluosilicic acid explains its 

 relation to silicic acid in the most satisfactory way. The substance 



* Am. Chem. J., iii. 189. 



