NUMBER OF ANNUAL GENERATIONS. 13 
destructive numbers is preceded by one or more scattering generations 
in the same field. 
Other things being equal, the worm must appear earliest in the south- 
ernmost latitudes, since extended observations on the appearance of 
Other insects show that there is retardation of from four to seven days 
with each degree of latitude northward. 13 
There is, in normal seasons of little injury to the crop, a similar re- 
tardation northward in the appearance of the Cotton Worm within the 
southern portion of the belt, corresponding in some measure with the 
growth and development of the plant; and it is a notable fact that the 
worm is seldom noticed and never in great numbers before the plant be- 
gins to bloom. What is generally under these circumstances called the 
first brood or "crop" has been preceded by at least one and often two 
generations Sparsely distributed over the fields. Vet in years when the 
worm abounds to a disastrous extent in the southern portion of the belt, 
its appearance in the northern or temporary portion cannot be counted 
on witli any certainty as to time, because it is always the result of mi- 
grations in the winged state, and these migrations may be more or less 
extended according to circumstances. Between tin' first appearance of 
the worm in the southern and northern portions of the belt there is, 
therefore, a marked ditt'crence ordinarily observable, it being in the Lit- 
ter much later and in far greater numbers. 
NUMBER OF ANNUAL GENERATIONS. 
The general impression and belief that prevails in the South is that 
there are, in those sections where the worm is most injurious, three 
broods or generations, or, as the planter puts it, "crops, '* each year. 
This Statement of the case has also been accepted as correct by most 
previous writers on the subject. 14 It is, however, essentially erroneous, 
so far as the southern portion of the cotton belt is concerned, as the 
earlier and later generations are not taken into account, but overlooked. 
The appearance of the first generation lias already been discussed, and 
it occurs during the latter part of April and in May m the more southern 
portions of the bett, and even in March in Florida. One generation 
follows another continuously from that time on, just so long as there are 
any leaves to be devoured, and we have, by protecting both plant and 
insect from frost, kept the moths ovipositing in the city of Washington 
all through November; while the worms, under like conditions, have 
hatched through the early part of December, matured, and spun up 
about Christmas time. 
Careful observations and experiments in 1871) in South Texas show 
that at least seven, and probably even more, annual generations are 
produced there. The first two generations are generally well separated, 
but, owing to the irregularity in egg laying and in individual develop- 
ment, the later generations so cross each other, that the insect may be 
