ACTION OF PHOSPHIDE OF SODIUM ON HALOID ETHERS, 205 
production of this salt would be an almost crucial test. We therefore proceeded 
with the supposed cacodyl as we had done with the oxide of tribenzyl-phosphine, 
and operating under exactly the same conditions, obtained a double salt with 
zinc iodide, which could not be distinguished rie that of the oxide, either in 
crystalline form or in composition. 
We have further verified the identity of the supposed cacodyl with oxide 
of tribenzyl-phosphine, by processes which we may consider along with the 
properties of that substance. 
Since writing the above, we have noticed that oxide of tribenzyl-phosphine 
has been obtained by F. FLEISsNER,* by the action of benzal chloride on iodide 
. of phosphonium. The results of FLEISSNER’s investigations on the properties of 
the oxide, so far as they go, are in accordance with our own. 
Oxide of Tribenzyl-Phosphine.—Subjoined are the results of the analysis of 
the oxide prepared by three different methods :-— 
I. Obtained as just described from the residues. 
II. and III. Obtained by the action of caustic baryta on chloride of tetra- 
benzyl-phosphonium. 
IV. Obtained during experiments on the action of sodium on chloride of 
tetrabenzyl-phosphonium (see p. 211). 
Obtained. Calculated for 
P(C,H;)30 
I. Il. Ill. IV. 
Carbon, . . ; Seb ise 79:2 788 78:3 78°75 
Hydrogen, . Ss) 6°8 6'8 67 6°56 
=~ 
Phosphorus, . : . 85 88 — — 8:4 9°68 
Oxygen, : : oo _ i 5°01 
100-00 
The three specimens were quite different in appearance. 
I. Crystallised in thick needles of great refractive power, and quite trans- 
parent. 
II. and III. In opaque plates of indefinite form. 
IV. In very bulky, silky needles. 
We could not at first reconcile ourselves to the belief that they were one 
and the same body. 
The following carefully conducted experiments, however, appear to prove 
beyond doubt that they were so :— 
Melting Point.—This was determined in the whee manner, by heating the 
carefully dried and pulverised substance in a capillary tube tied to a thermo- 
* Ber. d. deutsch. chem. Ges., xiii, 1665. 
VOL. XXX. PART I. 21 
