30 
capture to the laboratory. The adults, with the exception of 10 speci- 
mens killed at the time of capture, were kept alive as long as possible 
and dissected as soon as death was imminent. Some, in order that 
data might be obtained on their condition at the time of engorge- 
ment, were killed when captured and dissected within an hour. 
RECORDS OF DISSECTIONS. 
DEVELOPMENT OF THE OVARIES OF EEAEED NONFED ADULTS. 
In order to have a check on the experiments with engorged adults, 
32 females reared from isolated pupse were kept alive as long as possi- 
ble, without food, in bottles containing damp cloths. There were a 
number of males, reared from this same lot of pupse, at liberty within 
the breeding cage together with these females, and copulation prob- 
ably took place, but no definite statement can be made with regard to 
this point. Four females were killed and dissected within 6 hours 
after emergence. The remainder lived from a minimum period of 
14r| hours to a maximum of 119 hours. All these specimens on dissec- 
tion contained rather transparent ovaries of great potential develop- 
ment and full of round eggs (stage 1) (PL I, fig. 1, p. 4). In no 
instance did any one of these specimens develop the eggs within the 
ovaries as far as even the second stage. The maximum longevity of 
any nonfed reared specimen exceeded the maximum longevity of any 
specimen taken after engorgement by 52 hours. It would seem, 
therefore, that of these 32 reared specimens, there would have been 
1 or 2 which would have developed the eggs within the ovaries, if it 
had been in the natural course of events for them to do so. 
CONDITION OF THE OVAEIES OF FEMALES FOUND OVIPOSITING. 
Eleven females were taken actually engaged in oviposition on grass 
blades in streams. When these specimens were dissected, there was 
found in each stomach a certain amount of substance closely resem- 
bling the digested blood found in engorged specimens which had 
lived after engorgement for 40 hours or more previous to their dis- 
section. A chemical test for haematin was applied to the substance 
found in the stomachs of the females found ovipositing, in order to 
prove beyond doubt whether it contained animal blood or not, but, 
owing to the small quantity and the changes consequent to the process 
of digestion, it was impossible definitely to determine its character. 
Three specimens were taken from grass blades in a stream on which 
they had just alighted and had not begun to oviposit. These speci- 
mens were at once dissected, and their ovaries were found to contain 
from 250 to 300 fully developed eggs. Four specimens were taken 
from egg masses deposited in the same situation, which had evidently 
