THE GERMINATION OF COTTONSEED 1 
By Eben H. Toole, Physiologist , and Pearl L. Drummond, formerly Junior 
Seed Botanist , Seed Testing Laboratory , Bureau of Plant Industry, United 
States Department of Agriculture 2 
INTRODUCTION 
During 1921 and 1922, G. W. R. Davidson, of the Office of Crop Acclimatisa¬ 
tion and Adaptation Investigations of this Department, called to our attention 
the fact that our germination tests of cottonseed, especially of seed from Texas, 
gave very variable and often unaccountably low results. At about the same 
time the Texas Agricultural Experiment Station discussed with the Seed Testing 
Laboratory the difficulties of securing reliable results in making germination 
tests of cottonseed. We have attempted to analyze the situation and to locate 
the difficulties. Although all cottonseed are sensitive to conditions of germina¬ 
tion, a good sample will germinate promptly and vigorously under a compara¬ 
tively wide range of moisture and temperature conditions. Occasional samples, 
due to heating in storage or to other factors, contain many seeds with dead or 
weak embryos; such samples will not germinate vigorously under any conditions. 
On the other hand, a great many samples of cottonseed, when germinated by the 
standard method, 8 are conspicuous because of excessive mold and because of the 
u damping off” of the sprouts. This condition is most frequent in Texas seed 
of certain varieties. 
VARIATIONS IN GERMINATION RESULTS 
One of the great difficulties in germinating cottonseed from Texas has been 
the great variation in the results obtained from different tests of the same sample. 
Fourteen separate tests of a single sample by the standard method, all put in 
the germinator at the same time, gave results ranging from 34 per cent to 58 
per cent; with another sample, the results varied from 14 per cent to 46 per cent. 
These same samples, when tested in small boxes of soil in the germinator, pro¬ 
duced over 90 per cent of seedlings which were uniformly vigorous and fully as 
strong as those from samples which did not mold by the standard method. 
Great variations in the results of germination tests occur also when various meth¬ 
ods are used in making the tests, as is illustrated by the average germination of 
12 samples, all of one variety and received at the same time from a large seed 
firm. The average germination of the 12 samples was as follows: Standard 
method, 65.6 per cent; in soil in laboratory, 84.1 per cent; in sphagnum moss, 
85.1 per cent; in soil in a cool greenhouse, 72.9 per cent; in soil in a warm green¬ 
house, 92.3 per cent. 
i Received for publication March 17, 1924. 
8 The writers are indebted to W. L. Goss, G. W. R. Davidson, H. O. McNamara, W. W. Ballard, and 
B. M. King, of this Bureau, and A. H. Leidigh of the Texas Agricultural Experiment Station for aid in the 
field tests and for helpful suggestions during the progress of the work. 
8 As given in the “Rules for seed testing,” adopted by the Association of Official Seed Analysts of 
North America, 4 the dry seeds are put between folds of moist cotton flannel or of moist absorbent paper 
toweling and kept in a germinator at the usual 20-30° C. alternation of temperatures (30° C. for 6 to 8 
hours, and 20° C. for the remainder of the day). 
4 Association of Official Seed Analysts of North America, rules for seed testing. As 
recommended at its 10th annual meeting, 1917. Revised 1921. (Mimeographed). 
Journal of Agricultural Research, 
Washington, D. C. 
(285) 
Vol. XXVIII, No. 3 
Apr. 19,1924 
Key No. G-396 
