260 W. Gibbs and F. A. Genth 4 
It is not necessary to add chlorid of ammonium to the ammo* 
traces of impurity. It is not, however, necessary to use a pure 
chlorid of cobalt in preparing the chiorid of Prrpiareoodbell 
hour or two, almost all the original salt is decomposed, 7] 
given off in abundance during the boiling, while a lilac ae 
be mass 
supernatant liquid is to be poured off, and boiling water added 
to the insoluble portion. A brown-yellow or dark sherry Wine 
colored solution ‘is usually formed, which is again to be poured 
off, and the washing repeated till the liquid has a clear purple 
color. The red mass is then to be dissolved in boiling water, t? 
