The effects of forced practice in influencing handedness in the rat were first reported for animals subjected to cerebral injuries which had produced transfers (2). Milisen (1) later reported on a larger number of normal rats tested in a somewhat different food-reaching situation. The most systematic study has been reported by Wentworth (6), who produced forced practice with the non-preferred limb by means of an offset food dish. This arrangement has the food opening, into which the rat reaches, adjacent to one wall or partition of the cage so as to make it inconvenient for the animal to use one of his hands. Under these conditions, animals whose preference was unknown used the hand corresponding to the offset side with greater incidence than would be expected from known frequencies among unselected animals in the normal situation, where the dish is centered so as to favor neither hand. Since some animals, however, used the hand opposite to that expected from practice received in the offset situation, Wentworth concedes that this practice, while influential, is not the only determining factor in producing hand preferences. There is an assumption here that all animals used the more convenient hand in the offset situation, and Wentworth reports that he did not observe any use of the opposite hand in reaching. However, in this part of his study he did not take individual observations, nor count the number of reaches taken by his animals, but estimated them from Milisen's previous work. Casual observations did not reveal any opposite-hand reaches, but since the greater proportion of animals (approximately 90 per cent) gave correspondence to offset training, the exceptions could have been overlooked. In a similar offset situation we have occasionally seen animals make a very inconvenient reach with the opposite hand.Wentworth recognizes this difficulty but points out that tests to determine original preference necessarily give practice which may help to fixate such a preference (6, p. 73). In another phase of his study he used a minimum of such test reaches (ten reaches) to determine original preference, then forced the animals to practice with the opposite hand for 200 reaches in the offset situation. Of 56 animals so tested, 23 shifted completely, 18 were ambidextrous, and 15 returned to the originally preferred hand. The original tests and forced practice were run when the rats were 25 days old, and the retests were run 60 days later.The study of forced practice in animals with cerebral injury (2) made use of the normal food situation, which is not biased by being offset. The preferred hand was prevented from use by being bound with adhesive tape. When unbound, the limb is stiff for a day or two. Wentworth (6, p. 68) is critical of this method, dismissing it as unsatisfactory without stating precisely why. For certain purposes it may be less satisfactory than offset training, but this is on the assumption that the restraint of the limb and its subsequent stiffness will prevent its use in 184