It is remarkable that all of the 1: 1 compouncls crystallize in the distorted NiAs (MnP) structure rather than the NiAs structure itself. In the palladium and platinum systems where one might anticipate the NiAs structure no 1: 1 compounds were obtained.The diphosphides, diarsenicles, and diantimonides of ruthenium and osinium have the marcasite structure, while the corresponding compouncls of palladium and platinum have the pyrite structure. The rather complicated diffraction patterns of RhAs, and IrAs? are most certainly related, ancl bear a distinct similarity to the pattern of CoAs, (8). A strong low-order reflection in the CoAs2 pattern is split in the RhAs, and IrAs2 patterns; this leacls us to doubt the orthoi-hombic cell we proposed earlier for CoAs2. These diarsenides may prove to be isostruct~iral with the rhodi~im ancl iriclium diphosphides mentioned by Rundqvist.Since the tripl~ospl~icles and triantimonicles of cobalt, rhodium, and iridium are known to have the cubic CoAs:, structure, it is not surprising that RhAs:! and 1 1 -A~~ have the same structure. The list of compouncls crystallizing in the sltutterudite systenl now stands a t 1 I , ancl insofar as binary pnicticles are concerned, call be considered to be complete. With the exception of NIP, ancl PclI's they are formed only by the cobalt family of transition metals. Bism~itl~ides with this structure have not been observed.'I'hese systems arc no\v being st~icliecl in some detail.