HIBEKNATION AND DEVELOPMENT OF BOLL WEEVIL. 27 



developing- from them, or the fifth brood, emerged October 3, or in 

 23 days. In all cases this period of development includes to a greater 

 or less extent the feeding period, which must precede oviposition. 



Practically the same number of generations seems to occur at Col- 

 lege Station. Altho we have not determined this by rearing weevils 

 as rapidly as possible thruout the season, the field records show the 

 first three broods very distinctly, and during the fall there seem to be 

 three more broods between September 1 and the time of hibernation. 

 The number of broods after September 1 is very largely controlled 

 by the temperature and rainfall, which combine to afiect the food sup- 

 ply and the mortalit}^ of the developing weevils. It is probable, how- 

 ever, that in a late season seven broods might develop at College 

 Station, as the field records (see page 10, Table II, for 1903) show the 

 weevils (probablj^ of the sixth maximum brood) to have been freely 

 ovipositing early in November, and if frosts did not occur until late 

 in that month a large number of these would probably emerge. 



In southern Texas, Hunter and Hinds have shown that with the 

 average season of reproductive activity, extending from May 1 to 

 November 15, eight generations may be produced during the year. 

 Thus there is probably a difference of two broods in the maximum 

 number of generations between southern Texas and northern Texas, 

 but this difference will not materially affect either the injury by the 

 weevil or the number of weevils going into hibernation, as the weevils 

 seem to reproduce in practically the same length of time in northern 

 Texas as in southern Texas, the summer temperature being nearly the 

 same, and after the third brood they are sufficiently numerous in either 

 section to prevent the further fruiting of the crop. Nevertheless, 

 factors aside from a difference in the number of generations will prob- 

 ably cause less damage in northern Texas. 



By September 1 to 15 the weevils usually reach the maximum num- 

 ber and during the fall, as before observed, their number depends 

 primaril}^ on the food supply and weather conditions, so that the num- 

 ber going into hibernation depends on local conditions. The difference 

 in the amount of injury which the weevil may do in central or northern 

 and in southern Texas will depend therefore, first, upon the smaller 

 number of weevils hibernating successfully and, second, upon their 

 appearing later in the spring so that the cotton may be able to make a 

 more rapid growth in comparison with the development of the early 

 broods of the weevil under the high temperatures prevailing during 

 early summer in northern Texas than will be the case farther south in 

 late May. But it is not probable that the mere number of broods will 

 make any material difference in the amount of injury. The first three 

 broods seem to be well defined, as is clearly shown by the numerous 

 reports of our voluntar\^ observers and by the reports of the observers 

 of the crop service of the U. S. Weather Bureau. The amount of 



