PHOTOPERIODISM IN FISHES 663 



spawning of Esox vennicidatus (Lagler and Hubbs, 1943), suggests 

 a refractoriness ending in time for a stimulating day length-temperature 

 combination afforded by the unseasonable extension of warm days, to 

 take effect (see below). 



Before the perplexing heterogeneity of experimental results obtained 

 with Gasterosteus by the authors cited can be resolved into an intelligi- 

 ble pattern, the gonadal histology paralleling the onset and duration 

 of Baggerman's phases needs to be determined. In Gasterosteus, as in 

 Enneacanthus (Harrington, 1956), Perca (Turner, 1919), and appar- 

 ently Fimdidus (Matthews, 1938) and Gambiisia (Geiser, 1924), 

 spermatogenesis is completed far in advance of final maturation, so 

 that the quantified progress of ovogenesis affords the more likely means 

 of relating these phases to a measurable visible phenomenon. Day 

 length is without influence on spermatogenesis in at least one of these 

 forms (Burger, 1939, 1940; Matthews, 1939) contrary to the situation 

 in Phoxinus (Bullough, 1939) and Notropis (Harrington, 1957), 

 whereas there is no evidence so far of an ovarian cycle in this context 

 unaffected by changes in day length (cf. Atz, 1957). 



MODES OF DAY LENGTH MANIPULATION AND THEIR 

 RELATIVE EFFECTIVENESS 



The fish sex cycle has been influenced by several modes of day 

 length manipulation, among them abrupt imposition of long days, as 

 in most of the experiments so far cited, although in nature day length 

 changes are gradual and continuous, so that it is pertinent to search 

 for a common principle underlying the responses of fishes to the vari- 

 ous modes of manipulation of proved effectiveness. 



In the only autumn-spawning fish studied in this context, the brook 

 trout, Salvelinus fontmalis, sexual maturity was hastened by repro- 

 ducing a contracted version of the annual day length cycle such that 

 day length diminished earlier than in nature (Hoover and Hubbard, 

 1937). Since day lengths were gradually increased and then decreased, 

 the experiment failed to show whether the prior augmentation of day 

 lengths was an indispensable precondition of this precocious maturity 

 (Kleitman, 1949). Evidently it was not, for in subsequent experiments 

 (Hazard and Eddy, 1951 ) maturity was hastened one month (insufli- 



