236 EXPERIMENT STATION RECORD. [Vol. 40 



boll reaches a stage of immunity from weevil injury before it Is fully matured, 

 and that the length of the period before this stage is reached is apparently 

 more or less proportional to the total developmental period of the boll. Thus 

 while a fertile soil is more desirable for growing cotton under boll-weevil con- 

 ditions than a poor soil due to the larger amount of fruit set, nevertheless a 

 longer period is required on rich than on poor soil for the bolls to harden and 

 reach a stage immune to weevil injury. 



From a comparison of varietal differences in the fruiting processes it was 

 concluded that early commencement of flowering, usually associated with a 

 rapid rate of flowering, would be beneficial if the flowering could be maintained 

 In sufficient abundance for a longer period during the season unaccompanied by 

 a too high rate of shedding. However, early blooming varieties showed a 

 tendency to slow up In flowering activity relatively early in the - ■". and 

 also to lose much of the benefit of early flowering through excessive shedding. 

 These qualities appeared to be characteristic of Trice. Slmpklns, and I><> 

 Prolific Such varieties a> Wannamaker-Cleveland, on the other hand, though 

 blooming several days or even a week later than the earlier varieties, were 

 found to maintain nearly as high a rate of flower production per week after 



flowering began. These latter types also appeared to maintain a fair rate of 



flower production late in tin- season and with considerably less shedding tban 



the early types, a low rat Idlng, thus conserving a laru'e per- 



centage Of the early boll- garded as an important advantage under boll- 



weevil conditions. In the production of new types of cotton, therefore, it Is 

 deemed desirable not to overemphasise early flowering but rather to strive for 

 rapid (lower production combined with a low rate of early shedding even if 

 accompanied by Several days' delay In the appearance of the initial flowers. 



It is also thought to he desirable in general that a variety continue to grow 

 and fruit relatively late in the season in order to furnish squares with which 

 to occupy the weevils and thus protect the boils alread - long as possible. 



While it would be advantageous in relation to weevil injury, other things 



being equal, to have the same amount of cotton produced in a large Dumber 



of bolls rather than in a small number, the studies on flower production have 

 shown that the number of flowers produced by small boiled varieties was not 

 sufficient to counterbalance the lower lint yield of Individual fruits. The 



potential lint production of the big boll group thus considerably ex Hied that 



of the early small boll varieties. 



Observations on the relative resistance and susceptibility of the bolls of the 

 different varieties to weevil attack seemed to indicate that those varieties with 

 shorter boll periods got beyond damage earlier than those requiring a longer 

 period for development. Since the former were usually small boiled sorts mid 

 the latter usually large boiled, the large boiled varieties were at a disadvanti 

 in this respect. IIov • ome large boiled varieties with relatively high lint 



factors, for instance Wannamaker-Clevelaad, are Bald to possess a relatively 

 short boll period. 



The time at which cotton uses the most moisture. C. K. BfcClXLl \nt> [Jonr. 

 Amcr. Snr. Affron., 10 {1918), No. ' f . pp. t8i This paper, a contribution 



from the Georgia Experiment station, reports the results of experiments con- 

 ducted during I9ir> and I91fl to determine the water requirement of cotton. 

 plants were grown in potometers, maintained at different >'• satura- 



tion, and the quantity of water applied to individual plants per week during 

 July. August) and September and the number of blooms appearing each week 

 on each plant and on the plants studied under field conditions recorded in 

 tabular form. 



