12 REPORT 4, UNITED STATES ENTOMOLOGICAL COMMISSION. 
occupy twice that time. Taking the whole season through, however, 
the time from the egg of one generation to that of another will average 
about one month. 10 
TIME OF YEAR WHEN THE FIRST "WORMS APPEAR. 
Until the Cotton Worm investigation was begun, our knowledge as 
to the earliest appearance of the worms was not only vague, but mis- 
leading. The statement emphasized by Mr. Grote in the paper already 
referred to, namely, that the worm does not appear earlier than the latter 
part of June in the central portion of the cotton belt of Alabama and 
Georgia, very fairly echoes the prevailing popular belief on the subject; 
yet careful investigation shows the statement to be essentially erroneous. 
The date of earliest appearance varies with location, and largely with 
the curves of isochimal lines; 11 it also diners somewhat in different years 
in the same location, according as the season may be late or early ; and, 
lastly, it may differ to some extent in different parts of the same re- 
stricted locality, worms having been found just hatching in one place 
when, only a few miles distant, others were found nearly full grown. 
AVhile these modifying circumstances complicate consideration of the 
subject, it is easy to arrive at definite results by taking as a basis obser- 
vations made at a few particular points during the year. Hence we 
felt the importance of having such observations made during the spring 
of 1879 in South Texas and South Alabama at those places where the 
worm was reported to have appeared earliest in past years. As a 
result, the fact was fully established that the first worms of the season 
may, and do, in ordinary years, hatch from the middle of April to the 
middle of May in the southern portion of the cotton belt. And, more- 
over, all the facts showed that this season was a late one, for April 
frosts retarded the starting of cotton in those very sections of Alabama 
where the worms were first found ; while it is the unanimous opinion of 
planters in South Texas, where the worms were first noticed, that cotton 
was from two to three weeks later in 1879 than usual. Therefore, when 
in the spring of 1882 we found the worms of all sizes on rattoon cotton 
during the latter part of March, in South Georgia and Florida, we were 
not surprised, although this was fully six weeks earlier than they had 
ever before been noticed or recorded. 12 
The first worms are always comparatively few in number and in iso- 
lated spots. They are, therefore, easily overlooked by all who do not 
take particular pains to search for them. From such spots as centers 
the worms multiply and spread in subsequent generations, with greater 
or less rapidity, according as the conditions are favorable or otherwise. 
Such increase and spread may be confined to some part of a given county 
until the cotton is nearly ruined before the cotton in the rest of the 
county is affected. The worms will then first appear in the remainder 
much more suddenly and numerously than they did in the former, the 
parent moths migrating thereto in bevies. As a rule, however, the 
spread in the southern portion of the belt is gradual and the worm in 
