ECOLOGY OF SEED DORMANCY 161 



temperate-zone plants, which require 'stratification', namely, 

 pretreatment of the moist seed by cold^' ^. This restricts germi- 

 nation not only to post-winter conditions, thus minimizing the 

 dangers of frost killing, but also to regions with cold winters, 

 which most of these plants require for other developmental 

 phases (e.g. breaking of bud-dormancy, flowering). A similar 

 argument may be postulated for seeds the germination of which 

 is optimal at diurnally fluctuating temperatures^^* 27-29, 30-33 

 It is likely that such a requirement confines germination to 

 regions, seasons and soil depths where such fluctuations 

 occur^' 20' ^^. 



Control by specific light requirements may operate in germi- 

 nation in much the same way as it does in flowering, by providing 

 the process with a celestial clock. Short day seeds and long day 

 seeds have been described^-^' ^^. In addition, since the embryo is 

 heterotrophic, light-inhibited seeds may also be considered short 

 day seeds, while seeds which require even a flash of light are 

 comparable to long day seeds. Seeds of many hygrophytes 

 require light for germination^'^"^^. Here it is likely that the 

 requirement places a limit to the water-depth at which germi- 

 nation will occur 1. A similar mechanism may be operative in 

 plants which form forest undergrowth. 



Several problems present themselves in studies of this nature. 

 Two of these stand out because of their importance and the 

 danger of ignoring them. One is the fact that the vast majority 

 of our information on germination has been obtained by 

 experiments on or between filter papers, in Petri dishes or 

 germinators and using distilled water. Not only may some of the 

 data be a result of these experimental artifacts, but data may 

 have been overlooked. A case in point is that of lettuce seeds 

 which at 20^ appear light-insensitive when germinated in water, 

 but require light when germinated in solutions above a given 

 osmotic value^o j\^q other important consideration is funda- 

 mental to ecology. To what extent are the ecological implications 

 more than rationalization of phenomena which have little, if 

 anything, to do with the actual situation? It is dangerous as it is 



References p. 162 



