260 Dr. E. J. Mills on certain Cobaltamines. 



ordinary soft glass which contains lead is used, the brown- black 

 precipitate separates, for the most part, in beautiful lustrous 

 scales, and contains from 13 to 15 per cent, of silica; if hard 

 glass be substituted, the precipitate is destitute both of lustre 

 and of scaly appearance, and does not contain more than about 

 5 per cent, of that substance. Special comparative experiments 

 were made with aqueous ammonia of such a strength as the fil- 

 trates from these precipitates were found to possess ; but it was 

 found that there was no action on soft glass, and no diminution 

 in the ammoniacal standard. Some of the precipitates were di- 

 gested in sealed tubes with aqueous ammonia under like condi- 

 tions; but the glass was not at all corroded. Dicobaltic ter- 

 oxide and very strong aqueous sal-ammoniac, however, did 

 attack the tube slightly, and a little cobaltic chloride was formed 

 at the same time. It is the extent, not the nature, of such an 

 action which is surprising. The hydrates 



/3CoCl 3 -*(HO)'\5NH 3 , 

 CoCP-»(HO) w .6NH 3 , 



however ephemeral their existence under the circumstances, 

 ought to be much more energetic in their action upon glass than 

 potassic hydrate. 



(4) In section I. it is stated that the precipitates which have 

 just been alluded to, as well as dicobaltic teroxide (the product 

 of their complete oxidation), can be, to a great extent, trans- 

 formed into hexammoniochloride by digestion with sal-ammo- 

 niac and aqueous ammonia. In the case of dicobaltic teroxide, 

 we have the process of preparation of this valuable compound in 

 its most general form, the other methods given being manifestly 

 equivalent to this in their essential details. Either this process, 

 or the substitution in it of a chloride and a powerful oxidizing 

 agent for a peroxide, may prove to be useful in the future as a 

 means of obtaining polyammoniacal derivatives, if required. I 

 may mention, however, that I have not met with satisfactory 

 results on applying this method to nickel sulphate or zinc chlo- 

 ride, or by heating aqueous aniline or pyridine with sal-am- 

 moniac and dicobaltic teroxide. 



(5) When aqueous aniline, pyridine, or ethylamine (though 

 in this last case the experimental data have not been as nume- 

 rous as I could desire) is digested with cobaltic /3-pentammonio- 

 chloride under pressure, no analogue of the hexammoniochlo- 

 ride is formed, but the hexammoniochloride itself, together with 

 the usual products of the decomposition of the pentammonio- 

 chloride by heated water. As in these experiments no ammonia 

 was employed, it follows that the ammonia which gave rise to 

 the hexammoniochloride must have proceeded from the pentam- 



