The behavior of Ancylostoma tubaeforme (Zeder, 1800) larvae has been investigated in logarithmic thermal gradients of 20 C to 35 or 40 C over 1.5 cm. Larvae responded positively and when tracked in agar, the tracks were significantly straighter than those of the unstimulated controls. Tracks of larvae in the "2-heat experiment" showed a direct migration to the area between the heat sources and then a movement to either heat source, providing further evidence for a true thermotaxis. Distributional analyses have been developed as a rapid method for assaying tactic movements of nematodes when compared with "random walk" movements. The rate of movement of larvae responding up a steep gradient was more rapid than if warmed slowly at the same temperatures. Finally these results have prompted a hypothesis to explain the mechanism of thermopositive behavior of larvae of A. tubaeforme as a "behavioral overshoot," and the selective advantages are discussed. Among the conflicting reports of larval nematode orientation responses, the thermopositive movements of actively penetrating larvae of warm-blooded hosts have been considered one of the best substantiated (Payne, 1923; Wallace, 1961; Komiya and Yasuraska, 1966; Croll, 1970). Confusion and problems of interpretation still exist, however, because of the convection currents in some experimental designs, and the ambiquity between thermopositive and photopositive responses recorded using insufficiently cooled light sources (Parker and Haley, 1960). Surprisingly all reports of "thermotaxes" in nematodes have been based on distributional analyses following a period in a heat gradient. These designs have overlooked the possibility of aggregations occurring through kinetic changes in activity. Infective larvae have been tracked on agar and analyses of movement patterns attempted for Trichonema infective larvae (Croll, 1971) and for Ancylostoma tubaeforme (unpublished). From these and other tracks (Croll, 1969) it is known that arcs, caused by persistent asymmetrical biases in undulations, together with occasional reversal periods in which larvae move backwards for short distances, dominate larval movement patterns. The technique of accurately recording the movements of individuals is here applied to the behavior of A. tubaeforme larvae in thermal gradients. MATERIALS AND METHODS A. tubaeforme were cultured in moist cat feces at 30 C and tracked individually on 2% ion agar Received for publication 17 February 1972. (Croll, 1971). The conditions for tracking are so critical that we are confident that the water content of the agar, the nature of the surface film, and the texture of the surface were constant in each case. Larval movement patterns were also obtained by copying the positions of individuals on a grid which corresponded to one on the bottom of an agar-filled petri dish. Confidence can be placed on this technique, although it is less satisfactory than laboriously copying tracks inscribed in agar. Larvae were tracked at constant temperatures in Fisons' 50/G2 controlled-environment chambers, in 9-cm diameter agar plates held horizontally in the dark. When tracked individually larvae were removed at the end of the period. Because of idiosyncracies in larval movement patterns (Croll, 1971) and the tendency for individuals to follow a repeatable form of tracks, individuals which were tracked responding to thermal gradients were included in the "controls" which moved in nondirectional conditions. If the distribution of many larvae after a controlled period was examined, the plates were rapidly flamed over a Bunsen at the end of the experiments to kill the larvae without altering their distribution. A thermal gradient was maintained by placing a resistance wire in a glass tube into agar to a depth of 0.5 cm. The gradient on the agar surface was measured using a thermistor probe, until it was constant (about 20 min). All larval movements in gradients were observed in even, cooled, tungsten illumination of 5 to 10 c/ft2, passing perpendicularly to the larval movement, through the agar. Two heat sources were established by burying resistance wire in a "U tube" of 2 mm diameter, such that the heated elements passed upwards through the agar surface 2 cm apart. Those experiments testing distribution in directional light parallel to the agar surface used a "cooled" tungsten bulb of 100 c/ft2 intensity at th dge of the agar plate. The "straightness" of tracks has been quantified using: (1) tortuosity, or displacement/length, where displacement is the shortest distance between the start and end of the