It has long been recognised that the amount of ammonia in soils cannot be determined by distillation with a solution of caustic soda or potash since these strong alkalis slowly decompose the complex nitrogenous matter and evolve a coutinuous stream of ammonia. Boussingault showed more than fifty years ago how this difficulty could be obviated. By using magnesia in place of soda or potash the decomposition of such substances as urea, asparagin, and albumin was not great even on long boiling, but when the distillation was carried out under reduced pressure at 38°—40° there was no decomposition at all. Ammonium salts, on the other hand, were completely broken up in these circumstances. He used this low pressure method in his researches on the ammonia content of urine, but does not appear to have applied it to soils. Sufficiently accurate results could, he considered, be obtained by distillation at 100°, and this method was for many years generally adopted. It gave results varying according to the nature of the soil from 10 to 100 parts of nitrogen per million (·001 to ·01 per cent.), which may still be found quoted in some of the agricultural text books.