48 
Psyche 
[March 
in small denies and clusters isolated by several sorts of barriers, some 
of which would seem insignificant to a casual observer. It is fortunate 
that much of the Gunnison Basin is so far undisturbed by man and 
allows a base line for analysis of geographic variation in this species. 
Methods 
The basic requirement was to make a large enough sample to be 
reasonably accurate within the limited available time and yet not 
disturb the composition of the next generation. Therefore capture 
sampling methods were not used except for special purposes. In each 
locality we tried to count over 200 individuals on an area of 30 to 
50 acres. The sex ratio observed was typically greater than 10 males 
to one female due to the lubberly behavior of the females. We esti- 
mated the average density as 10 to 20 males per acre, so the sampled 
population was between 20% and 30% of the total male population 
of the sampled area. Each observer selected a different sector and 
walked a non-repeating path through likely areas of heavy grass- 
hopper concentration. As the insects made their escape flights or 
spontaneous social flights, the color of the wings was classed as 
either orange or yellow, since intermediate shades are difficult to 
distinguish from orange during flight. The landing place of each 
individual was noted and if an insect subsequently flew up within 
a foot of that area it was ignored, unless the color were the opposite 
shade. If there was any doubt as to whether an insect was an inde- 
pendent observation, it was not counted. Since the average density 
was about 10-20 per acre and the greatest density was about 60 per 
acre in isolated clusters, we had no difficulty in tracing individuals. 
One difficulty was noted early in the study; often the orange 
wings would not be visible because of shadow or peculiar angles and 
the yellow abdomen of the male would flash brightly instead. For 
this reason, any indefinite yellow-flashing individual would be forced 
to fly again. A second chance of error in sampling is the probability 
that the rarer color is counted more accurately than the dominant 
color. If a counted grasshopper should land near another of the 
same color and only the second one should fly up again, by our rules 
it would not be counted, whereas one of a different color would be 
counted. Therefore, the wing-color variant which is rare in the 
population could conceivably be of a lower frequency than actually 
recorded. A third difficulty encountered was the presence of another 
species of springbrooded oedipodine, X anthippus corallipes , whose 
wings typically flash lemon-yellow in flight. Pink-winged variants 
occur rarely in these populations. Experienced observers can easily 
