THE COTTON PLANT gI 
or only very occasionally cross-fertilised, until the 
writer showed in 1905 that from 5 to 10 per cent. 
of the ovules were habitually crossed under actual 
field conditions. Then it was easy to see why new 
cottons deteriorated or acclimatised themselves, why no 
new variety ever reached the market in a truly pure 
state, and why uniform cotton was unobtainable. Leake 
demonstrated independently that the same conclusions 
applied to Indian cottons, and drew the logical, but 
daring, deduction which we have just quoted. 
Without the assistance of photographs, sketches and 
graphs, it is hardly worth while to attempt any account 
of the life-story of a cotton plant, and the reader who is 
desirous of further information may perhaps be referred 
to the writer’s Development and Properties of Raw 
Cotton, but it might be of use to draw attention to a 
method of study which the writer has worked out to 
some degree of refinement for the study of Egyptian 
cotton, and by means of which some permanent results 
have been obtained, together with a fairly adequate 
comprehension of the manner in which any result 
was reached. This method is simply the obtaining of 
a continuous statistical record of the life-history, which 
can be plotted in graphic Curves of Plant-development, 
studying the crop—whether of a plot or of a country— 
as an Average Plant. Cardinal points are chosen for 
observation, the chief of these in the case of cotton 
being the number of flowers opening per plant per day, 
together with the height of the main stem and the 
ripening of the capsules or “ bolls.” The records can 
be carried to an incredible degree of precision by suit- 
