320 
Journal of Agricultural Research 
Vol. XXIV, No. 4 
at first. The plants which came up were counted daily and every four 
or five days their height was recorded also. At the end of 21 days 
they were measured for the last time, cut off at the surface of the ground, 
and dried to constant weight at 100® P. Data were thus secured on the 
days to germinate, the average height, and the dry weight of the plants. 
Most of the discussion in this paper, however, will be based on the dry 
weight which is perhaps the most reliable index of growth. 
Fig. I. —Diagram showing the arrangement of the tumblers in the preliminary experiments. A uniform 
concentration of sodium carbcmate was present in all ^e tumblers, and an equal total quantity of the 
sulphates was added to each in the relative proportion indicated by the shading. 
PREI.IMINARY EXPERIMENTS 
During the summer of 1919 some preliminary work was done on this 
problem. The technic of the experiment was essentially the same as 
that described above, except that no mulch was used and the preliminary 
mixing of the carbonate with the soil was not quite so thoroughly done. 
The tumblers were arranged in a triangular formation consisting of 66 
glasses. All contained a single concentration of sodium carbonate 
together with sodium sulphate, calcium sulphate, and sulphuric acid 
in the different concentrations represented in figure i. The total con¬ 
centration in parts per million of the added sulphates was always the 
