57 
(8) ; then no more until 1905, when a few issued May 24 (2), 29 (3), 
June 5 (1), 7 (1) ; then no more — observations ceasing August 8. In 
this case, the majority emerged near the middle of August and the 
minority waited until the next year. 
Taking heads collected still later in the season — August 26, 1904 — 
these gave imagines August 30 (a few), 31 (a few), September 1 
(few), 2 (few), 4, 5, 6, 7, 8, 9, 11, 12, 14, 15, 16, 17, 18, 19, 20, 21, 23, 
25, 26, 29, October 6 (one individual on each of these last dates, with 
two exceptions) ; then, in 1905, May 12 (5), 16 (many), 19 (many), 
20 (8), 23 (1), 26 (1), June 2(1), and no more up to September 1, 
when the daily observations were discontinued. 
Heads collected September 7 gave a few adults in September and 
October, but very many more in the following April and May. 
Heads collected September 16, October 3, 17, and 27, gave no 
adults until the following April, May, and June, when large numbers 
issued. 
In the last three instances most or all of the adults did not issue 
until the next year. Thus the time of emergence is delayed in propor- 
tion to the lateness of oviposition. 
These records are by no means exceptional ; they are typical ; those 
given here were selected from some twenty as being the most complete 
and most significant for present purposes. 
To summarize : The May and June adults lay their eggs in the 
first growth of second-year red clover, and most of the adults derived 
from these eggs appear in July and August, but some of them do not 
issue until the following May or June. The July and August adults 
lay their eggs in the second growth, and some of the adults from these 
appear during the same season ; the rest not until the following year. 
The later the eggs are laid the larger the proportion of individuals to 
lie over until the next year. The adults have two times of greatest 
abundance, falling near June 12 and August 12 respectively. This 
does not mean only two generations, however, but, instead, two lots of 
individuals. The first lot (June 12) is derived from eggs laid thruout 
the previous season. The second lot (August 12) has no accessions 
from the previous year, but is, nevertheless, composite in its nature, 
for it consists of adults of both the second generation and the third. 
I have traced as many as three generations per year, and there is 
clearly a possibility of four. The fact that some members of each gen- 
eration lie over until the next year complicates the study of the life 
history. 
At first, it seemed to me at least possible that the prolongation of 
the period of emergence might really be due to re-oviposition in the old 
clover heads, even tho the chalcids had been removed every day as fast 
as they came to light and no copulation had been observed. This as- 
sumption was dismissed, however, after being tested, for the seeds 
were all too dry and hard to be penetrated by the ovipositor or to serve 
as larval food. There remained only the possibility that some chal- 
cids had emerged soon enough after the clover heads were collected to 
