90 Stiles.—Concentration of the Nutrient Solution and the 
The second difficulty arises from the phenomenon of selective absorp¬ 
tion. All ions are not absorbed by the plant at the same rate ; the result 
is that not only the concentrations but the relative proportion of the con¬ 
stituents of the nutrient solution is also changing. 
In order to reduce the errors arising from these sources, seeds were used 
which were of as pure a strain as could be obtained, and which should there¬ 
fore have yielded plants as alike as possible. The seeds were germinated 
in clean sand, and young seedlings as much alike as possible were selected. 
The plants were grown singly in glass bottles of 1,200 c.c. capacity 
and were done in sets of ten. The corks used were coated with paraffin, 
and the solutions were changed every few (five to three) days, except in 
some instances where inquiry was made into the effect of not changing the 
nutrient solution. All the cultures in any one series were started on the 
same day and were also harvested and dried at the same time, so that the 
results are strictly comparable. 
Each plant was dried and weighed separately and the probable error 
of the mean of each set of ten results calculated, so that an idea of the 
significance of any differences in dry weight might be obtained. 
The nutrient solutions used were of four different concentrations, but 
the proportions of the contained salts were the same in each. Ordinary 
‘ pure 5 salts were used and a practically pure distilled water free from 
copper. The relative concentrations (i, •§•, y 1 ^, were the same as were 
used by Hall, Brenchley, and Underwood, but the actual proportion of salts 
was slightly different. The composition of the strongest solution was as 
follows : 
kno 3 
1 
CaS 0 4 , 2 H 2 0 
0-25 
MgS 0 4 , 7 H 2 0 
0-25 
kh 2 po 4 
°-2 5 
NaCl 
0-04 
Fe(N0,3)3. 6 H 2 0 
0*04 
Water 
1,000 
The Results. 
A preliminary series was grown during the early months of the year. 
The plant used was a Danish strain of Rye(Fej0,No.2), obtained for Professor 
Priestley by the kind offices of Professor Johannsen. Growth at this time 
of the year was slow, and consequently when the plants were removed they 
had not made very much growth. The actual results are as follows. The 
dry weight of each plant is given. 
