8 
which is quickly produced when the mixed solutions are 
heated. If such a decomposition indeed take place, we shall 
have, says Graham, a reaction, which, occurring in the soil, 
may lead to the ultimate production of those alkaline car- 
bonates required by plants for their nutrition. The steps 
in the process would be these : — decomposition of the alka- 
line chloride by calcium sulphate with production of alka- 
line sulphate; decomposition of the alkaline sulphate by 
lime, added to or originally present in the soil, with produc- 
tion of alkaline hydrate ; transformation of the alkaline 
hydrate into carbonate by the action of carbonates in the 
soil, or of carbonic acid in the water and air. 
The investigations of Graham and others have shown that 
the two last stages are readily accomplished. 
2. In order to determine whether the first stage in the 
series of transformations is or is not attainable, I mixed 
solutions of calcium sulphate and sodium chloride, and 
allowed the liquids to remain at the ordinary temperature 
of the air (15° to 18°) for several weeks. If decomposition 
had taken place the liquids would contain calcium sulphate, 
sodium chloride, calcium chloride, and sodium sulphate : the 
first of these salts is insoluble in ordinary alcohol : hence the 
addition of alcohol to the mixture would throw down the 
whole of the calcium sulphate existing as such, without 
otherwise influencing the state of equilibrium of the various 
salts in the liquid. In order to determine the amount of 
calcium sulphate which had undergone decomposition it was 
therefore only necessary to wash the precipitate produced 
by adding alcohol, with dilute spirit, and to ignite and 
weigh it. 
The following are my results : 
(a.) 100 cc. of CaS0 4 solution = 191*0 mgms. CaS0 4 mixed with 
10 co. of NaCl solution = 233 ’3 mgms. NaCl. 
Amount of CaS0 4 in solution after 28 days = 167’0 mgms. 
Hence amount of Na 2 S0 4 found = 25 mgms. 
