86 BOTANICAL GAZETTE [FEBRUARY _ 
of soil treatment is to be estimated by the result upon plant growth, 
it is essential that the roots of the test plants be living within the soil 
instead of being external to it. Fig. 2 shows at the left the soil mass 
from a paraflined wire pot, and at the right that from a pot with a 
non-adhesive wall. Wheat plants had grown in each pot for three 
weeks at the time the photograph was made. In the paraffined 
pot the plant roots were entirely restricted to the soil mass, while 
in the other pot the absorbing region of a majority of the roots was 
external to the soil and hence not indicating response to the treatment 
given to that soil. 
oo . ae 5” aa 
a 
Fic. 2.—Root distribution in paraffined wire pot and in ordinary pot; on the 
left is shown the contents of a paraffined wire pot, and on the right those of an ordinary 
pot. 
During the course of an experiment the pots were kept in situations 
where the conditions of heat and light were as nearly uniform as 
possible. 
The exact duration of the different experiments naturally varied 
somewhat on account of external factors, but was usually 18 to 21 
days. During very hot summer weather the period was often some- 
what shortened; the best conditions for growth in Washington, D. C., 
being found in the months of May and September. In order to 
eliminate errors which might unavoidably arise, as well as inherent 
differences in individual plants, five pots, each .containing six seed- 
lings, were used for each test; therefore, in recording the effect of any 
treatment, the total transpiration and green weight of the 30 plants 
distributed in five pots were taken. The total transpiration was de- 
termined by summing up the total losses of water from the pots dur- 
