662 EEPORT— 1887. 



of tlie bottle were evapor.ated to dryness on the water-bath in a platinum dish, 

 the residue treated with hot water, filtered, and washed. The filtrate was now 

 precipitated with argentic nitrate in the usual manner, and the precipitate, after 

 filtering, washing, and drying, was weighed. It was now reduced by heating in a 

 current of hydrogen. From these weights the quantities of chlorine and bromine 

 were calculated. 



During the distillation, decomposition occurred, and a large quantity of a sub- 

 stance containing silicon remained behind. 



The experiment was repeated, the mixture being warmed to about 45° for three 

 or four days and treated as in the previous experiment. On distillation, during 

 which decomposition again took place, it yielded the principal fraction between 

 120° to 125'^, which gave the following analysis : — 



Si 5-89 % 



CI 10-05 % 



Br 7-07 % 



From these numbers we get 



Si : CU or 1 : 2-02. 



The presence of Br here must be due to decomposition of the substance during 

 distillation, and must exist in the fraction as HBr, since bromine joined to O.jH^ 

 would not be so split oft" by NH^HO. 



The halogens were determined as described above. The silicon was estimated 

 by Polisa method,' which consists in heating the substance with strong K2SO4, 

 and oxidation with a solution of KMnO^. 



The method was modified by the use of sodium in a finely divided state. This 

 was prepared by melting the sodium under boiling toluene in a flask, and whilst 

 molten, corking the flask with a good cork and shaldng violently for a few seconds. 

 On allowing to cool, the toluene was poured off" and the sodium washed with dry 

 ether. 



As the chief fraction of the liquid obtained in the last experiment indicated 

 that only half the chlorine was removed from the silicon tetrachloride, and, more- 

 over, a large quantity of sodium was left intact, which probably resulted from 

 incompleteness of the reaction, only half the former quantity of sodium was there- 

 fore added. 



10 grs. SiCl^ (1 mol.) were mixed with 22 grs. CJTiBr^ (2 mols.), and this now 

 diluted with thrice its volume of dry ether ; to this 5'7 grs. (2 mols.) of finely 

 divided sodium were added, and a few drops of acetic ether. The reaction pro- 

 ceeded rapidly, and was completed on the water-bath. On allowing to cool, the 

 liquid, which was of a light brown colour, was poured off", and the greater part of 

 the ether distilled. As in the former experiments, decomposition took place on 

 distillation ; it was therefore determined to examine the liquid without fractiona- 

 tion at all. The remaining fluid was placed in vacuo over sulphuric acid and 

 allowed to remain there four days. The resulting liquid was now dark brown in 

 colour, thick, and fumed in the air. On exposure to air it became coated with a 

 solid substance, and for this reason was removed to small bulbs for analysis, as 

 quickly as possible. 



Si determination. By Polis's method. 

 0-59 gr. gave 0-11375 SiO., = 8-997 % Si 

 Found. Caic. for. (C.,H^Br)„ SiCl, 



8-997 % Si. '8-88 % Si. 



The follo-ndng equation shows the reaction that occurs: — 



CH,Br 



2C,H,Br, + SiCl, + 2Na, = ois^siCl, + 2XaCl + 2XaBr 



CH„Br 



' Ber. DcidscTi. Chem. Ges. xix. 1024. 



