Dodge continues his account of technique and results in his work on compensatory eye-movements. Any attempt to measure eye-reaction compensatory to bodily rotation involves the difficulties of securing sudden rotation of the body mass, and the recording of head- and eye-movements.In this study the S sat on a turn-table which was set in rotation by means of rubber springs held in tension by E until the moment of release. All accessory apparatus rotated with the S and head-and eye-lines were recorded. The difficult problem of obtaining an adequate rotation record was solved by placing a concave mirror on a shaft rising from an excentric axis. The shaft was kept motionless while the turn-table revolved about it by a friction belt from a pulley on a tripod base to a pulley of the same diameter on the excentric shaft.The records revealed that: (1) the average latent time of reflex compensatory eye-movements was three or four times shorter than the latency of saccadic eye-reactions (2) there was a ςo-ordinate' compensatory eye-movement, compensatory to voluntary head-movement. The latency of this type of eye-movement was extraordinarily short. The latencies of both of these types of compensatory eye-movement were much less than the latency of ocular pursuit movements which begin after a natural reaction time and (3) pursuit movements were accompanied by a series of inaccurate approximations to pursuit, whereas reflex compensations showed less hesitation in onset. Co-ordinate movements of compensation were distinguished from the above types in this respect by an astonishing smoothness. From Psych Bulletin 19:03:00163. (PsycINFO Database Record (c) 2006 APA, all rights reserved)