202 PROFESSOR LETTS AND N. COLLIE ON THE 
action of phosphorus on sodium ; for instance, a phosphide analogous in com- 
position to the liquid phosphide of hydrogen, viz., Na,P,, or even to the 
solid phosphide, viz., Na,P,. If these bodies were indeed formed, they might 
give by their action on chloride of benzyl analogous benzyl compounds, and of 
these the phosphorised cacodyl of benzyl, 7.e. (C,;H,),P,, would possess a 
great interest. . 
We therefore determined to submit the residues to a very careful examina- 
tion, the result of which we have now to communicate. 
On boiling them with chloroform, the whole of the organic ingredients 
appeared to be dissolved, and the residue consisted apparently of amorphous 
phosphorus only. The chloroform extract was distilled to dryness, and there 
remained a solid gummy, slightly crystalline mass. This was boiled with 
alcohol, when the greater part dissolved, leaving about one-third undissolved 
in the form of a brown amorphous solid. The brown alcoholic extract was 
filtered and mixed with about one and a half times its volume of boiling 
water; the mixture was then boiled and filtered from a brown resinous 
substance which had been precipitated. The filtered solution deposited on 
cooling a considerable quantity of a crystalline substance, though in a very 
impure condition. 
We thus split up the residues into three portions— 
1. Soluble in chloroform only. 
2. Soluble in chloroform and alcohol. 
3. Soluble in chloroform, alcohol, and a mixture of alcohol and water. 
We may here observe that as the portion of the residues No. 1 was not 
very inviting, in later experiments the residues were at once extracted with 
alcohol, so as to obtain only portions 2 and 3. 
Examination of Residue soluble in Chloroform, Alcohol, and Alcohol and 
Water.—This portion of the residue was distinctly crystalline, but highly 
charged with a brown resinous substance, which threatened at first to render 
its purification difficult. 
At first we attempted to recrystallise it from alcohol, but although well- 
defined crystals were easily obtained, the colouring matter adhered persistently; 
animal charcoal was then boiled with the alcoholic solution of some of the 
crystals which were only slightly coloured, and eventually by this process a 
colourless, crystalline, and highly refractive substance was obtained. 
But we soon found that the crude and strongly coloured residue could be 
purified in a far easier and simpler manner. On treating it with ether or 
bisulphide of carbon, the colouring matter was at once dissolved, but the 
crystalline substance was quite insoluble, so that by simple washing with either 
of these liquids, an almost colourless residue was obtained, which only required 
to be recrystallised once or twice from hot alcohol. The crystals thus obtained 
