CONTROL OF ANNUAL GONADAL CYCLES 725 



EFFECT OF INTERRUPTED LIGHT IN PHOTOPERIODICALLY 

 INDUCED TESTICULAR DEVELOPMENT 



It has been demonstrated in domestic ducks (Benoit, 1936a), 

 Stunms vulgaris (Burger, Bissonnette, and Doolittle, 1942), Colinus 

 virginicmiis (Kirkpatrick and Leopold, 1952; Kirkpatrick, 1955), 

 Jiinco hyemalis (Jenner and Engels, 1952), and Zonotrichia leuco- 

 phrys gambelii (Farner, Mewaldt, and Irving, 1953a,b) that a given 

 total duration of light becomes more effective in stimulating testicular 

 development if it is broken into shorter periods. In general the 

 effectiveness appears to increase as the duration of the intervening 

 dark periods is reduced. In all these experiments, although com- 

 parisons of rates cannot always be made accurately, it appears that 

 the rate of response rarely, if ever, exceeds that of a continuous light 

 period equal in duration to the experimental light periods plus the 

 intervening dark period (s), a relationship of fundamental importance 

 in the interpretation of the effects of interrupted light. 



The recognition of the logarithmic nature of testicular growth in 

 Zonotrichia leucophrys gambelii has permitted a more precise exam- 

 ination of the interrupted-light phenomenon by comparisons of the 

 values for k of equation (2) for a wide variety of photoperiodic treat- 

 ments (Farner and Wilson, 1957a, and unpublished data). In Fig. 3 

 it is readily apparent that 6 hr of light (incandescent lamps, supra- 

 maximal intensity) in equally spaced 50-min photoperiods causes 

 the same rate of testicular development as a single daily photoperiod 

 of about 12 hr. These investigations also show clearly that the effec- 

 tiveness of very short photoperiods is a function of the intervening 

 dark period. Thus 2-sec flashes with 15-min intervening dark periods 

 are ineffective, whereas 2-sec flashes whh either 58-sec or 8-sec dark 

 intervals give near maximum values for k. Similarly, with intervening 

 dark periods of 15 min, 2-sec light flashes, as noted above, are 

 ineffective, whereas 20-sec light periods give near maximum values 

 of A:. 



These and other data (Farner and Wilson, unpublished) lead us 

 to the hypothesis that somewhere in the response mechanism there 

 is a rate-limiting reaction which, during the light period, produces 



