Variation in thoracic color and the role of color in reproduction were studied in the field in one population of the damselfly, Argia apicalis. Mature males and females were individually marked, and observed near noon on 47 consecutive days, and hourly on three. Males occurred in two color phases: bright blue, and gray-black, neither of which could be positively associated with age or mating. Change was not a single step from blue to finally dark or vice versa but often involved intervening changes in both directions with a maximum of eight in 12 days. Dark was the more temporary condition. Females occurred in three color phases: brown, turquoise, gray-black. As with males, no one phase could be positively associated with age or mating, and multi-directional change occurred after sexual maturity.
Dead pinned individuals, modified or not, were presented to living males who: (1) advanced sexually toward motionless models, (2) discriminated intact females from males, (3) reacted sexually more frequently to brown than to the other normal female colors, (4) responded sexually to a female thorax and one wing almost as frequently as to a normal female, (5) accurately discriminated a female thorax and one wing from that of a male. The blue-tipped male abdomen aided but was not indispensable for sex recognition. Dorsal and lateral thoracic color were equally important and elicited male sexual response in their entirety rather than by particular pattern. Sex discrimination broke down when the normal thoracic color was obliterated with paint regardless of its color, but a thorax painted white was an "over-optimal" attraction suggesting the importance of light intensity.