of the Cacodyl Series. 385 



The composition of this radical is therefore 



Calculated. 1. 2. 



Carbon 4- equivalents 23-1.5 22-30 22-23 

 Hydrogen 6 ... 5-67 5-48 5-33 



Arsenic 2 ... 71-18 71-29 71- 



Loss and Oxygen . . . 0*00 0-93 1-44 



100- 100- 100- 



The trifling difference between the quantities found and the 

 calculated quantities, arises probably from the impossibility of 

 obtaining this compound free from oxygen. If the results 

 obtained" are reckoned in the 100 parts, without taking notice 

 of the oxygen, the carbon and the hydrogen agree still closer. 

 The quantity of arsenic on the contrary appears rather too 

 much. The facility with which cacodyl can be separated 

 from its compounds by simple substances, renders it very 

 probable that the oxide might be also reduced by means of 

 carbon as well as hydrogen, upon the application of a higher 

 temperature. Dumas's analysis, as well as my previous one, of 

 the liquor of Cadet, renders this supposition nearlj' certain, 

 and fully explains the cause of our arriving at different results. 

 Dumas found, as I did also, in my first experiments, a con- 

 stant excess of arsenic, carbon, and hydrogen, which is ac- 

 counted for by the impurity of the oxide of cacod3'l. 



There was no difficulty in ascertaining the density of the 

 vapour of the liquid, as the temperature at which it is decom- 

 posed is considerably higher than its boiling point. 



Substance 0'2500 gramme. 



Volume measured .... 55-98 Cbr. 



Temperature 200° C. 



Barometer 328-5 lines. 



Column of oil 38 lines. 



Col. of mere, at 200° C. 44-5 lines. 

 This gives the density of 7-101, which agrees as nearly as 

 could be expected with the calculated density, viz. — 

 4 volumes of vapour of carbon . 3-371 



12 ... hydrogen 0-825 



2 ... vapour of arsenic . 10*367 



14-563 -=-2 = 7-281. 



The difference of 0-18 in the result obtained is fully ac- 

 counted for by the tension of the mercury vapour in the ba- 

 rometer at the temperature of 200° C. 



The agreement of both the analysis and density of the va- 

 pour with the respective calculated quantities is a matter of 

 considerable interest. Berzelius has shown that when a cer- 

 tain density of a gaseous organical radical is assumed, the 



Phil. Mag. S. 3. Vol. 20. No. 132. Mai/ 1842. 2 D 



