POLISHING METAL SECTIONS 855 equal to zero. On the other hand, the failure of Moissan' and of Ruff' to obtain any evidence for the existence of free NH, a t very low temperatures is strong evidence to the contrary. Furthermore, it has been shown by the writer3 that the alkali and alkaline-earth metals exist in mercury, even in dilute solution, not in the form of single atoms, but in that of compounds of the general formula hlHg,', containing only one atom of the amalgamated metal to the molecule. Ammonium amalgam behaves analogously to these, and we are justified in concluding that it also is a solution of a compound of the general formula hf.Hg,,, in mercury.T h e solution of the coinpound (NH,)Hg,, is not stable, however, except at very low temperatures. I t decomposes slowly at oD, more rapidly a t room temperature, into mercury, ammonia, and hydrogen. T h e gases become entangled in the mass and give rise to the characteristic inflation, which, however, is not a property of the original compound, but merely an accidental phenonietion accompanying its decomposition.