24 
taneously at any time after the middle of May. The winter of 1890-'91 
was unusually mild in Lousiana, and the spring proportionately earlier. 
Hence the above dates may not be average or normal, and, in any case, 
are intended to have only a local application. 
The foregoing discussion is based upon observations made in northern 
Louisiana and Mississippi. In northern Mississippi the evidences of a 
portion of the last brood hibernating as imagos are more meager and 
less conclusive. In 'Arkansas the reports of observers and the time 
of greatest depredation seem to point conclusively to the fact that all 
of the last brood hibernate in the pupal state, while from -the fact that 
the spring is later than further south, their habits of hibernation are 
more constant, the first brood issuing more evenly and all the broods 
being better defined. The fall season is also more severe, if not earlier, 
and hence only the five broods occur in the cotton-producing portion 
of the State. In the remainder perhaps only four full broods occur, 
with an incomplete fifth one. On the contrary, in southern Texas the 
winters are mild and the spring comes much earlier than in the cotton 
region of Louisiana or northern Texas. From constant communication 
with cotton-planters and other observers in southern Texas, it was 
determined that there could be no doubt about the hibernation of a 
considerable portion of the last brood as imagos. These appeared and 
began depositing earlier than at Shreveport, certainly producing six 
distinct broods and a partial seventh by the close of the season. Those 
issuing from pupfe in spring produced six well-defiued broods. 
Failure to take into consideration these geographical and meteoro- 
logical differences over so vast a territory as the culture of cotton 
occupies has resulted in great confusion and much controversy among 
cotton-planters as to the number of broods and the times of their appear- 
ance. The truth probably is that each is correct for his own district. 
The determination of the time of appearance of the several broods of 
moths and when their egg-laying is most abundant is a matter of 
great importance in intelligently managing the trap-crop method for 
protecting the cotton, and will be further discussed hereinafter. 
PARASITES. 
On June 15 specimens of Pteromalus puparum obtained from Pieris 
rapw in great numbers, were placed with a large Boll Worm upon earth 
in a wide-mouthed bottle. As both males and females of the parasites 
had been x^laced in the bottle, some were seen copulating later. The par- 
asites frequently alighted on the back of the larva. The worm, open- 
ing its jaws, would quickly and violently throw its head and the anterior 
part of its body around to the point where the insect sat, and often cap- 
tured it. This was not merely accidental, for the process was often re- 
peated and a micro was nearly always captured. Once in the jaws of the 
larva, the parasite was quickly eaten. Before pupation, June 19, tfhe larva 
had in this manner eaten about thirty of the forty or fifty parasitic in- 
