REPRODUCTION AND DISPERSAL 



903 



into flowers, and again nearly all of the primordia may develop into bul- 

 bils. Usually the bulbils germinate while still attached to the inflores- 

 cence, but they are readily detachable, and they continue growing if 

 they fall into favorable situations. Similar bulbils occur somewhat 

 regularly in certain alpine plants, such as Polygonum viviparum, Poa 

 alpina, and Saxifraga. The cause of such bulbil formation is not clearly 

 known, though in the grasses, seeds 

 rather than bulbils appear to be pro- 

 duced, when the plants are grown in 

 dry conditions or in media that are poor 

 in nitrogen. That bulbil formation in 

 alpine plants probably is a reaction to 

 alpine conditions is indicated by the 

 fact that Gagea plants taken from low- 

 land to alpine habitats produce bulbils 

 rather than flowers in the first season 

 of culture. In the onion this habit 

 affords the plant the apparent advantage 

 of an additional method of reproduc- 

 tion, supplementing reproduction by 

 seeds and by subterranean bulbs ; in 

 alpine plants bulbil formation may be 

 more certain than seed maturation, 

 owing to the shortness of the season. 



Fig. 1203. — An umbel of the 

 wild garlic (Allium canadense), in 

 which there are only a few flowers 

 (/), their place being taken largely 

 by bulbils (A), which readily propa- 

 gate the species vegetatively ; />, 

 perianth segments; s, spathe. 



The factors determining the development 

 and form of the organs of vegetative repro- 

 duction have been sufficiently considered elsewhere. It may be noted merely 

 that in specialized organs, such as tubers and bulbs, xerophytic conditions may 

 favor development, much as in other kinds of reproductive organs. For ordinary 

 vegetative reproduction, however, particularly where it is indistinguishable from 

 the usual phenomena of growth, mesophytic or hydrophytic conditions are more 

 favorable. 



In conclusion, it is now clear that external factors play an important 

 part in determining the variation of reproductive organs. How great 

 this part may be, and what the precise external factors are, remain as 

 yet in large part unknown. It cannot suffice to explain phenomena by 

 such terms as " bad nutrition " or " xerophytic conditions," though to 

 attribute them to such a cause-complex is vastly more satisfactory than 

 to refer them to inherent or internal causes. 



