Mason . — Growth and Abscission in Sea Island Cotton . 465 
amount of variation and the apparent absence of any significant relationship, 
it has not been reproduced. 
The environmental factors have been displaced five days to the left in 
each of the figures. Such a procedure, of course, assumes that a constant 
interval of approximately five days intervenes between the incidence of the 
causative factor and the completion of abscission. It was found, however, 
that by doing this a reasonably close correspondence could be traced between 
certain external factors and the major waves of boll-shedding. Ewing’s 
studies, too, it may be noted, indicated that five to six days represented 
Fig. 6. Daily rates of boll-shedding, rainfall, evaporation, and evaporation due to solar 
radiation. Gp. I. 
under natural conditions the mean interval between the stimulus and the 
abscission response, while Lloyd’s work in Alabama pointed to an approxi- 
mately similar period. 
Reference to Fig. 4 will show that the two main waves of shedding 
(commencing October 2nd and 10th respectively) were both preceded by 
periods of heavy daytime rain and low rates of evaporation. To the minor 
waves, commencing October 26, no significance can be attached, for the 
probability exists that they would have been eliminated had the population 
been larger. In Fig. 7, for instance, which shows the results for the 
749 plants of the second group, these minor irregularities are almost 
