New YorkK AGRICULTURAL EXPERIMENT STATION. 831 
ment is illustrated and partly described in Bulletin 146.% It is 
of iron-frame construction and is so arranged that the condi- 
tions of light, heat and ventilation may be kept remarkably uni- 
form throughout. The portion used for the plants is separated 
from the walls by walks. The plants were grown in boxes ar- 
ranged on benches as indicated in the diagram opposite. The 
numbers in this diagram correspond to the soil treatment num- 
bers given in Table IV. 
THE BOXES. 
In testing the action of particular factors upon plants it is 
essential that all conditions be under the best possible control 
so that the other factors which are not being tested shall be uni- 
form. At the same time it is desirable that the conditions of the 
experiment conform to ordinary horticultural practice so far 
as this can be done without lessening the reliability of the test. 
It was decided that these conditions could best be fulfilled in 
these experiments by growing the plants in small wooden boxes. 
These have an advantage over benches or sections of benches 
in that plants receiving similar treatment may be distributed in 
various locations in the house. With such an arrangement there 
is a tendency to equalize in the average results any inequalities 
which may exist in the environment in different locations in the 
house. Moreover boxes may be easily handled for weighing or 
photographing. The boxes were made 15x 15x88 inches inside. 
They were not filled quite to the top and some space at the 
bottom was occupied with gravel for drainage so that the soil 
in each box was about 7 inches deep, which is not far from the 
depth of soil ordinarily found on greenhouse benches. The 
boxes easily accommodate four lettuce plants, one near each 
corner. 
THE SOILS. 
Both a medium clay loam and a very light sandy loam were 
used in these experiments. Each particular treatment as regards 

*®N. Y. Agr. Exp. Sta. Bul. 146 :162 and Pl. II; Ann. Rept. 1898 : 471 and 
Pl. XLIV, 
