988 
Journal of Agricultural Research 
Vol. V, No. 21 
ditions are the principal factor in the production of these forms, yet 
different, perhaps more severe, conditions are needed for each succeed¬ 
ing generation. Also, the generation itself becomes of more and more 
importance, till in the eighteenth (first produced on September 30) every 
experiment contains some sexes, while in earlier generations batches of 
young containing only parthenogenetic females were produced after that 
date. This latter point is emphasized by the fact that in the nineteenth 
generation only sexes appeared, while in the earlier generations some 
viviparous insects were produced as late as were any of the insects in the 
nineteenth. 
It should be stated that the first sexes, in the open, were observed 
about September 15. These were partially grown. By September 22 
adult and nearly full-grown males and females were abundant, indi¬ 
cating that these forms were produced at least as early as the 6th of 
September. 
PERCENTAGE OP MALES TO FEMALES 
Notes were not made in every case of the numbers of males and 
females in an experiment, but the records of 71 experiments in which 
such figures were kept give an average of 11 per cent of males in a total 
of 350 insects. This is above the true average, since many experiments 
contained “many females and no males/’ and such records have not 
been included. In only four experiments did the males outnumber the 
females, and in these experiments the greatest number of sexes raised 
was six. 
length op nymphal life 
The period covered by the nymphal life of this form was considerably 
longer than that covered by the same stages of viviparous females, 
although there were only four nymphal stages, as in those forms. The 
average period for the immature stages was 20.6 days, the range being 
from 16 to 36 days. It was impossible to obtain satisfactory data as to 
the divisions of this period occupied by each stage, as in the majority of 
the oviparous females'the normal rate of growth was considerably 
deranged by cold spells. Such conditions would greatly retard the 
development of the insect, with the result that the particular stage in 
which the insect passed through such temperatures was lengthened in 
comparison with the other stages. Thus, one experiment might show 
the first to be the longest stage, while in another the longest stage might 
be the third. In the case of oviparous females bom early in September, 
the first three stages occupied about the same amount of time as the 
entire nymphal period of the viviparous females, while the last stage 
continued for about 6 days. Later in the fall it was impossible to make a 
comparison. The males require the same amount of time for complete 
development as do the females and the length of the nymphal period is 
affected by climatic conditions in exactly the same manner for both sexes. 
