462 Journal of Agricultural Research voi. xvui, No. 9 



Only two flies of the fourth summer brood were obtained. These 

 emerged in cage 87 on September 24 and 25. The small number was 

 due to the rapid dwindling of the broods in cages, as explained under 

 "Methods" p. 459. When a possible failure to get the fourth brood in 

 lineal descent was anticipated, a supplementary series of cages was 

 started, stocked with flies obtained by sweeping a bluegrass lawn. One 

 of these, cage 92, gave adults, and they correspond very well with those 

 of cage 87, the first fly appearing on the same date in both. Cage 92 

 gave two adults on September 26 and one, the last of the season, on 

 October 3. 



Cages started the last of September with flies swept from the lawn and 

 with the few reared flies gave no results, no eggs being obtained. The 

 weather was cool, and the flies were almost continuously dormant. After- 

 thought would suggest that in spring and fall the cages need a good deal 

 of direct sunlight. No indications, however, of a fifth summer brood 

 were observed. Flies emerging in September probably live longer than 

 those of midsummer, having long dormant periods in cool weather; so 

 they merely lay eggs on winter wheat in October for the winter brood. 



The record here given covers too few flies, and those kept under too 

 uniform conditions, to exclude the possibility that a portion of the rep- 

 resentatives of this species, under natural conditions, might have a brood 

 more or less. It does indicate, however, that four is the normal number 

 of summer broods; and, allowing for the effect of a slight retardation 

 in the cages in spring and fall, it is probable that five broods in the sea- 

 son will occur oftener than three. 



LENGTH OF INSTARS 



The number of days from adult to adult in each of the summer broods 

 is shown by Table III. 



In this table there were 8 individuals in brood II and 3 in brood 

 III whose parents emerged two days apart, the intermediate day 

 being taken as the date for both. One of these cases in brood III gave 

 the minimum record of 22 days. The actual minimum period for the 

 season was for the male of this pair. He was 21 days from emergence 

 when his offspring emerged; but since his mate was 23 days old the 

 number 22 was recorded as the average. The period covered in this 

 case for the male was July 21 to August 11, in which there were two 

 hot waves separated by a few^Bomewhat cooler days — on the whole, an 

 excessively hot period. 



The table shows an average period from adult to adult in the first 

 summer brood of 49.7 days for 35 individuals; for the second brood, 

 30.3 days for 41 individuals; for the third brood, 28.5 days for 21 indi- 

 viduals; and for the fourth brood, 45.5 days for 2 individuals. 



