Temperature of Sublimation. 5 



The decomposition is, it will be seen, for the most part 

 •effected within the temperature range of group III. It is 

 probable that the liberation of As below this is referable to 

 the presence of the molecule As 2 S 3 in subordinate amount, 

 accounting for the S present and the slight excess of As. 

 The composition is, however, essentially that of a diarsenide, 

 and the break up of this molecule must be referred to the 

 temperature range of group III. 



The position in the scale of temperature of the intervals of 

 stability and instability appear to be only roughly definable 

 even on successive experiments on the same powdered material. 

 The temperature at which an interval of decomposition or 

 instability comes to an end depends to some extent, ap- 

 parently, on the duration of heating at the lower tempera- 

 tures embraced by the interval. Thus the temperature 320° 

 which marks the conclusion of an interval of instability, as 

 recorded above, on another experiment was found to be some 

 30 degrees lower ; the heating at the lower part of the 

 interval having been much more prolonged. The tempera- 

 ture of commencing instability, 230°, was, however, the same 

 in each experiment. Similarly the second interval of insta- 

 bility began at temperatures which were found to be fairly 

 concordant. The upper successive nodes and internodes 

 seemed less well defined, but their position on the scale of 

 temperature cannot be investigated without special pre- 

 cautions to secure uniform increase of temperature and 

 continuous observation of sublimation. 



With regard to the cause of these successive intervals of 

 stability and instability, it seems probable that the interval 

 230°-3'20°, which covers that of group II., indicates the 

 presence of the As 2 S 3 molecule, the existence of which is, 

 also, to be inferred from the analysis. Over the next 

 interval the diarsenide molecule breaks up. From its rela- 

 tively large amount it is not probable that this constituent is 

 in any way generated by rearrangements attending the evo- 

 lution of arsenic at the lower temperature. It is, however, 

 quite possible that rearrangements attending the break up of 

 the diarsenide molecule are responsible for the cessation 

 of instability, the succeeding interval of stability, and for 

 the decomposition which follows this interval. The firm 

 retention of a considerable part of the volatile element till 

 a very high temperature is attained, strongly suggests the 

 formation of a compound or alloy of the residual arsenic 

 with the stable element ; this compound possessing a very 

 high decree of stability-. 



