Bayley — On Complex Oxides of Cobalt and Nickel. 107 



Per Cent. 







Theory for NigOn, 9 H2O. 





Found. 



grams. 





grains. 



74-02 



(1) 



74-15 





(2) 



74-29 



MO 



Summary. 



Under the influence of the hypochlorite solution nickel and cobalt 

 form the oxides NisOs and C03O5. On boiling the liquid containing 

 CogOg it loses oxygen, and passes to the form CoijOig, intermediate be- 

 tween C02O3 and C03O5. Under similar circumstances, M3O5 appears 

 to decompose without forming stable lower oxides, although it is pro- 

 bable from the results that the f onnation of NioOa is a stage in the pro- 

 cess. The oxides C03O5 and C012O19 appear to be stable at a low red 

 heat ; they are distinguished from C03O4 by a slight difference of colour. 

 C03O3 has the following hydrates : — 



C03O5, 4 H3O (dried over H.SOi), 

 C03O5, 3 H2O (dried at lOO^C), 

 C03O5, 2 HoO (di-ied at 138T), 

 and probably, 



C03O5, HoO (dried at SOO^C). 



MsOs decomposes while still moist when its precipitating liquid is 

 removed by washing. In the one experiment which was made, the 

 resulting dried compound agreed closely with the formula NisOn, 

 9 H2O. I have found that, when C03O5, 4 HoO is treated with cold 

 dilute nitric acid, part is dissolved with evolution of oxygen, and that 

 part remains insoluble. I hope, in a future Paper, to give the results 

 of some similar experiments undertaken for the purpose of determining 

 the proximate constitution of these oxides. 



It may at first sight appear that the formula CojoOig is inadmissible 

 on account of its complexity, but as the iodine method clearly shows 

 that the oxide is exactly intermediate between C03O5 and C00O3, and as 

 the formula C012O19 is the simplest formula for such an oxide, it would 

 seem that we must accept it, especially when we consider the tendency 

 of cobalt to form compounds vieing in complexity with many of the 

 products of organic chemistry. It may be that the application of the 

 iodine method to the examination of the oxides of other metals would 

 lead to the acceptance of formulae more complex than those now admit- 

 ted. The two oxides of cobalt described in this Paper, and indeed 

 other oxides of cobalt, may be represented graphically by rings some- 



