470 Mason. — Groivth and Abscission in Sea Island Cotton . 
variability in the daily growth-rates of the individual plants and the 
percentage of buds and bolls shed in the same period were correlated or 
not. The smallness of the population and the very limited amount of 
shedding which occurred while growth was in progress must render such 
a procedure very dubious. Nor is it possible to obtain a single index 
which will express the amount of variation occurring in the growth-rate on 
consecutive days, for the coefficients of variability, it will be clear, would 
also express the amount of variation over more extended periods. 
The Growth-rate of the Main Axis and the Subterranean 
Environment. 
Though the results presented in the previous section seemed to render 
it improbable that the subterranean environment could play any considerable 
part in causing the growth-retardation which occurred during periods of 
daytime rain and low evaporation, it seemed none the less desirable to 
attempt to place the matter on an experimental basis. The growth-rates 
of four groups, each of ten plants, were accordingly determined daily. The 
plants were not quite twelve weeks of age when the measurements recorded 
in Fig. 9 were undertaken. With a view to avoiding the decline in the 
growth-rate which attends fruit development, all the bolls were pruned off 
some days previously and the flowers removed daily as they were produced. 
Drains about one foot in depth were made on each side of the third and 
fourth groups in order to ensure against the oxygen-supplying power of the 
soil being markedly diminished during periods of excessive rainfall. Inas- 
much, however, as rain is a saturated solution of oxygen and the soil of 
St. Vincent is extremely permeable, the possibility of such a contingency 
must be regarded as remote. To the first group approximately 10,000 c.c. 
of water, to which 2 c.c. of a 3 per cent, solution of hydrogen peroxide 
were added, were applied daily to the soil round each of the plants from 
October 17 until the end of the experiment. On October 16 a sheet of 
white linoleum extending about three feet on each side of the plants was 
placed over the surface of the soil in the neighbourhood of the fourth 
group. The space between the linoleum and the plants was sealed with 
wax ; the subterranean environment of this group was thus isolated from the 
direct effects of heavy rain. 
Inspection of the graphs in Fig. 9 reveals the remarkable fact that 
growth was apparently entirely suspended on October 9 in every group. 
The individual records showed an extreme range of from — o*8 to i*o cm. 
The measurements on the 9th were, however, made after midday, and not, 
owing to the torrential rain, which fell in the early part of the day, at 8 a.m. 
as usual. The net result of this delayed measurement may have been to 
diminish somewhat the growth credited to the 9th, and to have augmented 
that of the 8th. 
