The actual home range of almost every individual studied, even of those individuals for which the most data were available, probably differed at least slightly from the observed or estimated home range. One adult female, for example, was captured six times in two years within a radius of approximately 50 feet. Another female was found 2780 feet from her last point of capture. These last two records were regarded as unusual; when they were grouped with records of the 44 individuals mentioned above, the average radius of home range for the entire group was much larger (327 feet). [Illustration: FIG. 29. The movements of a gravid female of _T. o. ornata_ in the southern ravine area at the Damm Farm in a period of ten days in June, 1956. Her movements were, for the most part, in and around several ravines (shown on map by broken lines) where she was searching for a nesting site. For explanation of symbols see legend for Fig. 28.] Homing Behavior Gould (1957) reported that 22 of 43 _T. carolina_ moved in a homeward direction when they were released in open fields up to 5.8 miles from their original points of capture. Turtles oriented themselves by the sun; homeward headings were inaccurate or lacking on overcast days and, light reflected from a mirror caused turtles to alter their courses. Seven of ten turtles released more than 150 miles from home headed in directions that corresponded most nearly to the headings last taken (at release-points near home base) and did not necessarily correspond to the direction of home. Gould's studies point out that box turtles perhaps practice a kind of "solar navigation." His work raises the question of whether the movements of box turtles are guided by the sighting of local landmarks or whether such landmarks alter the course of movement only when acting as barriers.