Many attempts have been made to prepare nitro-glycerine explosives capable of withstanding comparatively low temperatures without freezing, but no satisfactory solution of the problem has been found. Among the substances that have been proposed and used with more or less success, are nitro- benzene, nitro-toluene, di-nitro-mono-chlorhydrine, solid nitro derivatives of toluene,[A] are stated to lower the freezing point of nitro-glycerine to -20 deg.C. without altering its sensitiveness and stability. The subject has been investigated by S. Nauckhoff,[B] who states that nitroglycerine can be cooled to temperatures (-40 deg. to -50 deg. C.) much below its true freezing point, without solidifying, by the addition of various substances. When cooled by means of a mixture of solid carbon, dioxide, and ether, it sets to a glassy mass, without any perceptible crystallisation. The mass when warmed to 0 deg.C. first rapidly liquefies and then begins to crystallise. The true freezing point of pure nitro- glycerine was found to be 12.3 deg.C. The technical product, owing to the presence of di-nitro-glycerine, freezes at 10.5 deg. C. According to Raoult's law, the lowering of the freezing point caused by _m_ grms. of a substance with the molecular weight M, when dissolved in 100 grms. of the solvent, is expressed by the formula: [Delta] = E(_m_/M), where E is a constant characteristic for the solvent in question. The value of E for nitro- glycerine was found to be 70.5 when calculated, according to Van't Hoff's formula, from the melting point and the latent heat of fusion of the substance. Determinations of the lowering of the freezing point of nitro- glycerine by additions of benzene, nitro-benzene, di-nitro-benzene, tri- nitro-benzene, p.-nitro-toluene, o.-nitro-toluene, di-nitro-toluene, naphthalene, nitro-naphthalene, di-nitro-naphthalene, ethyl acetate, ethyl nitrate, and methyl alcohol, gave results agreeing fairly well with Raoult's formula, except in the case of methyl alcohol, for which the calculated lowering of the freezing point was greater than that observed, probably owing to the formation of complex molecules in the solution. The results show that, in general, the capacity of a substance to lower the freezing point of nitro-glycerine depends, not upon its freezing point, or its chemical composition or constitution, but upon its molecular weight. Nauckhoff states that a suitable substance for dissolving in nitro- glycerine, in order to lower the freezing point of the latter, must have a relatively low molecular weight, must not appreciably diminish the explosive power and stability of the explosive, and must not be easily volatile at relatively high atmospheric temperatures; it should, if possible, be a solvent of nitro-cellulose, and in every case must not have a prejudicial influence on the gelatinisation of the nitro-cellulose. [Footnote A: Eng. Pat. 25,797, November 1904.] [Footnote B: _Z. Angew. Chem._, 1905, 18, 11-22, 53-60.]