342 REPORT OF THE COMMISSIONER OF AGRICULTURE. 



brood are to be seen in great numbers at nigiit sucking the nectar of 

 cotton, cassia, and cow-pea. 



The flfth brood begins early in September, and is also confined to cotton. 

 In ail but the most southern portions of the cotton-belt this brood ap- 

 pears normally to be the last, its chrysalides living through the winter 

 in their underground cells. With an exceptionally fine season it seems 

 probable that there may be another brood, but upon this point we have, 

 as yet, no evidence. 



These remarks upon the number of broods are based upon observa- 

 tions made the present year in Centi'al Alabama, and the following facts 

 must be taken into consideration : that the observations were limited 

 geographically to a single point, central, it is true, but were unconfirmed 

 by observations from other points. Moreover, 1879 was by no means a 

 bad worm year. From opposite extremes of the cotton belt we should 

 expect to find variation in opposing directions from this as an average. 

 In years when the worms were very numerous we should expect to find 

 them infesting cotton at a period earlier than that which we have 

 designated, and where com is not grown in the vicinity, they probably 

 feed upon cotton from the first appearance of the flower-buds. These 

 points will account for the fact of the frequent early reports of the rav- 

 age of the boll-worm in cotton. 



The same difficulty also arises in ascertaining the precise number of 

 broods of the boll-worm that was found with the cotton- worm. Some 

 moths issuing from winter quarters later than others, or failing so soon 

 to find a suitable i)lace of deposit for their eggs, will lay their eggs later 

 than others. Some larvae, moreover, may, by surrounding circumstan- 

 ces, fail to develop as fast as others. These and other points combined 

 start an irregularity of the broods, the tendency of which is to contin- 

 ually increase rather than to dimhnsh, until in the later generations 

 upon cotton we may find them in all stages at once — eggs, larvae of all 

 sizes, chrysalides, and moths. The number and relative appearances ot 

 the broods normally, however, we believe to be that which we have given. 



The boll- worm disappears in the fall before the cotton-worm does. 

 Mr. Gr. W. Smith-Vaniz, of Canton, Miss., gathered eggs from one of 

 which a larva hatched August 30. It became a chrysalis September 22, 

 and passed the winter in this state, issuing avS a moth May 14. Another 

 brood of the cotton caterpillars was reared after this boll- worm went 

 into winter quarters. 



In his Third Missouri Entomological Eeport, p. 107, Professor Eiley 

 makes the statement : 



Most of the moths issne in the fall and hihemate as snch, bnt some of them pass the 

 •winter in the chrysrJJs state and do not issue till the following spring. I liavo known 

 them to issue, in this latitude (36^° N.), after the 1st of November, vi hen uo fioat ha^ 

 previously occurred. 



It may be true that Reliothia occasionally hibernates as a moth. No 

 instance of such hibernation has, however, come under our notice, nor 

 do we find any other statement of this fact than this of Professor lliley's, 

 just quoted. It is certain that the insect normally hibernates in the 

 chrysalis state, and that if a hibernating moth is found it is an excep- 

 tional occurrence. 



Many of the noctuidse hibernate as moths, and some, as, for instance, 

 the army- worm of the north {lIcUoj>hila imipuncta, Haworth), are sup- 

 posed to winter either in the moth or chrysalis state. The latter point 

 is not yet definitely settled, however, and even if it were it would sim- 

 ply create a ])recedent not necessarily a probability, in favor of a dual 

 hibernation of Helioihis. 



