10 BULLETIN 828, U. S. DEPARTMENT OF AGRICULTURE. 
portions of the curves made show the same relations as in the instances where com- 
plete records were kept. In general, the earlier planted fields developed their max- 
imum wilt prevalence before Jhe later planted fields, but with a few notable excep- 
tions. Thus four out of six cucurbit fields planted between April 25 and May 1 de- 
veloped their maximum wilt between June 23 and 30, while four out of five fields 
planted after the first of May (mostly between May 15 and 20) developed their max- 
imum wilt about the middle of July. Of the exceptions, one field planted April 25 
and one planted May 15, each gave their maximum wilt on July 5. Reference to the 
notes in these cases showed the striped-beetle maximum in both fields to have come 
early in June. Furthermore, one cucumber field planted April 20 to 25 had its max- 
imum beetle prevalence June 12, and its maximum wilt came July 12; another cu- 
cumber field planted May 21, with beetle maximum between June 15 and 20, had its 
wilt maximum July 21. While wilt prevalence usually reached its maximum first 
in the earliest planted fields, there were these instances where it did not, and in these 
exceptional cases it was found the beetles had been late in making their appearance. 
Thus again the wilt prevalence has followed the striped-beetle prevalence rather 
than the time of planting or the weather conditions. 
Finally, during the season of 1917, careful counts were kept of beetle and wilt 
prevalence (fig. 4) in our experimental field near Tuxedo, Md. Here the hibernating 
brood of striped cucumber beetles reached its maximum numbers June 25, while 
the wilt in this field of 1^ acres reached its greatest prevalence July 21. A period of 
heavy rains preceded and followed the time of maximum wilt prevalence, and rain 
was fairly well distributed throughout the season. About August 10, the cucumber 
plants reached maturity and died. The field was replanted, but only about one- 
fifth of the seeds germinated. The young of the summer brood of striped beetles fed 
freely on these scattered seedlings, but no wilt was found. The young of a fall brood 
of striped cucumber beetles began to appear about the middle of September, but at 
this time all our cucurbits were either ripe or dead, so that their relation to the wilt 
was not observed. No attempt was made to determine the number of broods of the 12- 
spotted cucumber beetles, but the curve shows their general prevalence throughout 
the season. Except for short periods between the broods of the striped beetles, they 
were at no time present in as large numbers as the striped species. 
As will readily be seen from the preceding statements and from a 
comparison of the three sets of graphs (figs, 2 to 4), the wilt curves 
always bear a definite relation to the beetle curves, approximately 
one month occurring between the maxima of beetle and wilt curves. 
Wilt prevalence does not in these cases show any relation to time of 
planting, humidity, or rainfall, and little or none to temperature. 
About one month appears in all these sets of curves between the 
maxima of beetle and of wilt prevalence. Yet cucumber plants 
usually wilt within less than two weeks after they are inoculated, 
and the results of a spray test (cf. p. 39) in one of the fields where 
these beetle and wilt records (fig. 3) were kept show that the great- 
est number of infections actually did take place within the preceding 
2-weeks period. Why, then, are the maxima of the beetle and wilt 
curves so far apart ? Collating all our experimental data and obser- 
vations concerning beetle and wilt relations, the explanation appears 
to be that early in the spring only a very small percentage of the 
beetles are wilt carriers. As the season advances and larger and 
larger numbers of cucurbit plants become wilted a correspondingly 
