43 



and that of bromide (153C.) a difference which amounts to 

 94=3 x 31i Kopp derives the formulae 



SiCl 3 and SiBr.j, 



as representing the atomic constitution of the chloride and the 

 bromide of silicium, and he accordingly fixes the equivalent of 

 silicium at 21 '3. 



In order, however, to prove the general validity of Kopp's obser- 

 vations, it was necessary to re-examine the boiling-points of cor- 

 responding chlorine and bromine compounds which exhibited dis- 

 crepancies, and to extend the inquiry to as great a number of new 

 compounds as possible. 



Mr. Francis Baldwin Duppa has, at my suggestion, undertaken an 

 investigation of this subject, and has already obtained some valuable 

 results, whieh I beg to communicate to the Royal Society. 



The bromine-compound of titanium was unknown. Mr. Duppa 

 has produced this substance by passing a current of bromine over an 

 intimate mixture of pure titanic acid and carbon. The reaction 

 takes place at a bright red heat, and furnishes a brown liquid, which 

 solidifies in the receiver to a crystalline mass. Distilled with an 

 excess of mercury, which removes any free bromine that may be 

 present, the bromide of- titanium presents itself as an amber-yellow 

 compound, exhibiting a magnificent crystalline structure ; it attracts 

 moisture with the greatest avidity, and is converted into titanic and 

 hydrobromic acid. Bromide of titanium has a specific gravity of 

 2' 6. The fusing-point was found, 39. The boiling-point was 

 examined by Mr. Duppa with a considerable quantity of substance, 

 the purity of which had been ascertained by analysis. It was 

 observed to be 230 C. The boiling-point of the chloride of tita- 

 nium, as observed by Dumas, and confirmed by Mr. Duppa, is 135. 

 The difference, 230 135=95 = 3 x31|, is exactly the same as that 

 observed between the boiling-points of chloride and bromide of 

 silicium. 



This -observation furnishes an additional support to the analogy 

 of silicium and titanium, while it points unequivocally to the 

 formulae 



TiCl 8 and TiBr 3 , 



as representing the atomic constitution of these two compounds. 



