We found that in our marmosets, the eyes showed diurnal fluctuations in axial length, extending this phenomenon to a primate model of eye growth. In chicks, eyes elongate more during the day than during the night,
7 8 9 whereas in rabbits, the reverse is true.
10 In normal (untreated) eyes of both these species, the time of maximal elongation approximately coincides with the time of highest IOP, implicating a possible influence of the rhythm in IOP on the rhythm in axial length. Indeed, IOP has long been postulated to play a role in ocular development and ocular growth as a source of inflationary pressure (see, for example, Refs.
36 37 38 ). In the marmosets, however, this phase correspondence held true only in the adolescent animals, with their slower-growing eyes increasing in length during the night, when IOP was highest. The faster-growing eyes of juveniles, on the other hand, increased in length during the day, when IOP was lower. This age-related difference in the phases of the rhythm in axial length is not associated with a concurrent age-related phase difference in the rhythm in IOP that might account for this
(Fig. 7B) . These findings support the notion that IOP is not the only (or the main) influence on axial length, in agreement with several other studies showing a noncorrespondence in the phases of these two rhythms.
10 11 In rabbits, for instance, the rhythms in IOP and axial length are only approximately in phase: IOP peaks early in the dark phase,
14 15 whereas axial length peaks late in the dark phase.
10 Furthermore, in form-deprived myopic chick eyes, the rhythm in IOP is no longer synchronized to the light-dark cycle, yet the phase of the rhythm in axial elongation does not change accordingly.
11