10 TOXIC SUBSTANCE EXCRETED BY PLAN'S. 
plumule was passed upwards. In the latter cases the plant was 
adjusted so that the seed was in position just above the surface 
of the water in the bottle. 
The seedlings were very carefully chosen from several 
hundreds grown in crushed quartz, so as to be as nearly equal 
among themselves as possible. The state at which the seedlings 
are best suited to the purpose in hand was found by repeated 
experiment to be as follows :— 
(1) Cotton when the spread of the cotyledons is 14 inches. 
(2) Sorghum when the first and second leaves are equal 
(both about 1 inch long). 
(3) Cajanus when the ‘spread’ of the first leaves is 4 
inches. 
(4) Sesamum as soon as the cotyledons have assumed a 
horizontal position. 
(5) Wheat as for sorghum. 
(6) Gram when the first three leaves have expanded. 
The strength of the solutions had been so arranged (by 
allowing to evaporate) that no plant (of the size indicated) would 
grow in any of them for more than about ten days. The time 
between transplanting into the bottles and the times of com- 
mencement of wilting and of complete drying up were carefully 
recorded. The bottles were also weighed every morning and in 
some cases several times a day to find the amount of transpira- 
tion, this amount having been proved (Bulletin No. 28, Bureau of 
Soils, U. S. A.) to be a measure of the increase of the plant in 
dry weight. 
In all cases two, and in some cases as many as six bottles were 
treated in the same way, both as regards the solution they con- 
tained and the crops grown in them. It was found that with 
careful selection of seedlings of each size the difference between 
duplicates either in time of withering or in loss by transpiration 
was extremely small. The observations here recorded are in all 
cases the mean of the total number of bottles sown in the partic- 
ular manner indicated. They are set forth in the following 
tabular statement :— 
