140 
Psyche 
[Vol. 90 
and 39 mm long were found subsequently. The females (n = 81) 
ranged from 12-29 mm and 19-247 mg with the biggest female 2Vi 
times as long and 13 times as heavy as the smallest. At all lengths 
females were heavier than males, and they increased in weight faster 
with length than did the males (Fig. 3). 
A frequency histogram of the lengths of males (n = 101) and 
females (n = 128) measured June 9 is shown in Fig. 4. Mean male 
length ±S.D. was 25.91 ±7.21; mean female length ±S.D. was 20.97 
±4.35. 
Five females were dissected. Each had two ovarioles and 3 or 4 
large, yolked eggs. The length of the largest yolked egg increased 
monotonically with female length, from a 1.3 mm egg m a 13 mm 
female to a 2. 1 mm egg in a 29 mm female. 
Size ami Aggressive Success 
In aggressive encounters between females the female that fled was 
deemed the loser. The winners by this criterion were larger in 14 of 
14 contests involving weevils of unequal length (p = .0001, sign test). 
Even if four additional encounters involving females of equal length 
were conservatively counted as victories for the smaller weevil, the 
winners were still significantly more likely to be the larger (p = 
.0154). 
In male encounters the winner was considered to be the male that 
remained by the female. Here again the larger weevil was signifi- 
cantly more likely to win (p < .005, sign test). Defending males (the 
ones originally with the female) were not significantly more likely to 
win encounters than intruding males (p .18). 
The relative size of the rivals was also a factor in the occurrence of 
the appendage-fencing contest. An analysis of the differences in 
length between the rivals in five encounters in which the contest 
occurred and sixteen encounters in which it did not, showed that 
rivals using the contest were significantly more similar in length (p = 
.002, Mann-Whitney U test). The mean ±S.D. difference in length 
for rivals using the contest was 1.8 ±2.0 mm; for rivals not using it, 
8. 1 ±7.4 mm. 
One effect an intruding male may have, whether or not he wins 
the female, is to shorten the duration of the defending male’s copu- 
lation. Uninterrupted copulations lasted a mean ±S.D. of 82.4 
±48.7 sec, with 2/3 of the copulations lasting between 40 and 90 sec. 
Copulations interrupted by rivals, however, lasted 31.0 ±15.1 sec 
(p = .026, Mann-Whitney U test). 
