14 



found in all stages in the same field at one and the same time. Yet the 

 succession of broods may be recognized by the condition of the bulk of 

 the specimens in the field or by confining the insects for the better 

 watching of them. 



The first generation, as we have seen, is confined to spots. The sec- 

 ond generation is more dispersed, but still restricted to areas in the 

 vicinity of the hibernating centers. The third generation of worms 

 may become, under favoring conditions, not only widespread but dis- 

 astrous, and the moths produced from them so numerous that they 

 acquire the migrating habit, ifbis generation appears in South Texas 

 during the latter part of June, and in South Alabama and Georgia 

 somewhat later. This is usually the supposed first brood in those sec- 

 tions, i, e,, the first which attracts general attention. The subsequent 

 generations naturally become more and more widespread, and the in- 

 crease during July and August, 1880, was i^oticeable in the face of mete- 

 orological conditions unfavorable thereto. The worms during these 

 months will appear in those fields in which they did not appear earlier, 

 as on sandy, elevated prairies, soils or lands where the growth of the 

 plant was retarded from late planting, overflowing, the injury of the 

 plant-lice, or whatever cause. If the weather be favorable, this August 

 generation will, when unmolested, carry ruin in its wake. Did one gen- 

 eration follow another in the natural ratio of multiplication, such is the 

 fecundity of the species that there would be no hope of profitably culti- 

 vating cotton. Fortunately for man, some of the earlier generations 

 are liable to be so effectually kept in check by natural enemies and 

 other adverse influences that they become innocuous and frequently es- 

 cape notice. This fact Was strikingly illustrated in May, 1879, in Col- 

 orado and Lavaca Counties, in Texas, where the second generation, 

 which hatched in sufficient numbers in most fields to create alarm, never- 

 theless vanished before its enemies so completely that it was difficult, a 

 little later, to find a perfect chrysalis. 



That this second generation may exceptionally become very injurious is 

 shown from records, to the effect that in the early part of June of the same 

 year, w hile the cotton stem was yet tender, whole plantations in the low bot> 

 torn lands of Louisiana have been eaten to the ground j but that it more 

 often proves harmless is probable for various reasons. The plant-lice, 

 which are apt to be very numerous on the very young cotton, partially 

 disappear before their natural enemies by the time this second brood 

 of worms is developing. The ants, which were previously supplied 

 with food by the plant-lice, have now multiplied, and are forced, by the 

 decrease of the aphides, to seek other food. They are consequently 

 more effectual in destroying the young worms. All the other enemies 

 of the worm are also more active during the month of June, and gre- 

 garious birds, like the blackbirds and ricebirds, are very common during 

 that month, but generally leave the fields later. 



In the northern portion of the cotton belt the number of annual gen- 



