Calcium chloride, exposed to a stream of ammonia gas, rapidly absorbs it, and forms the compound CaCl₂.8NH_g. It would appear that this compound is one of calcamine, Ca(NH₂)₉, with 2HCI, with which six molecules of ammonia are associated in some manner resembling that in which water of crystallisation is associated in salts containing it. Thus we have CaCl₉.6H₂O ; and Ca(NH₃)₂Cl₂.6NH₃ has an analogous formula. Zinc and cadmium form similar compounds, and other salts may be obtained from the appropriate salts of the metals; thus, by saturating zinc sulphate with ammonia, the compound Zn(NH₃)₂SO₄.H₂O separates in crystals. Again, with aluminium, A1(NH₃)₃C1₃ has been prepared; and dyad iron, manganese, and nickel yield somewhat similar compounds. Such bodies must be regarded as salts of ammonium, 'in which a metal has taken the place of one atom of hydrogen in each molecule of ammonium; a dyad metal replacing two metals in two molecules of ammonium, a triad three, and so on.

The state of such compounds in solution is probably that of "double salts," alluded to on pp. 10 and 161. While some of them are decomposed by water into ammonia and the salt of the metal, others resist that decomposition, and are ionised into complex groups, analogous to the platini-chloride or the silicifluoride group. Thus, while it is probable that the compound of ammonia with calcium chloride in solution contains as ions Ca, -CI, NH₄> and -OH, together with non-ionised NH₄OH and molecular NH₃, the fact that zinc hydroxide, precipitated by addition of ammonium hydroxide to a solution of the chloride, is re-dissolved by further addition of ammonia, is doubtless to be explained by the formation of the complex ion Zn(NH₃)₀, which is soluble in water. But this does not exclude the presence of the usual ions, Zn, -CI, NH₄, and -OH, which doubtless co-exist with those already mentioned. In some cases, the stability of the complex ions is much greater than in that mentioned, and of this some instances will be given.