1903] Biological notes on Canadian species of Viola. 157 



pubescens, glabella and orbiculata ought to be reckoned of the same 

 group as V. Canadensis, but in these species as well as in the 

 more southern V. hastata, the above-ground stems are readily 

 seen to have developed directly from the rhizome and in the axils 

 of leaves, pertaining to the terminal bud. In other words, their 

 rhizome represents a monopodium just as typical as the one 

 described above, with the only difference that in the one case only 

 flowers develop, while flower-bearing stems develop in the other. 

 These flower-bearing stems die down to the ground without leav- 

 ing any basal buds for reproduction ; this is secured by the 

 terminal bud of the under-ground main-axis. 



In V. pedata the rhizome is constantly vertical ; it is rather 

 short, but quite thick, and the primary root persists for about 

 two years. In V. papilionacea, affinis, etc., the rhizome is hori- 

 zontal, somewhat longer, but the internodes are barely visible. 

 The thickness of the rhizome is in these species (B) as well as in 

 V. pedata due to the swollen bases of the petioles, as we have 

 described in a previously published paper. ^ In V. prvnulcBJolia 

 and its allies (E) the rhizome is quite slender, and these species 

 are characterized by their more or less profuse development of 

 long, very slender, subterranean stolons, each of which becomes 

 terminated by a rosette of leaves like the mother-shoot, and in 

 which the same monopodial branching takes place. 



The peculiar instance of both caulescent and acaulescent 

 flowers developed upon the same rhizome is illustrated by V, 

 rfitundifolia, as described above. However, our material wae 

 rather scant of this species, thus we were unable to make out 

 whether the cleistogamic flowers are borne on stems that are 

 aerial or subterranean under normal conditions. 



In V. sarmentosa we have a type which is very different from 

 all the others on account of its sarmentosc habit. The stolons 

 are, as it appears from our dried material, always above ground, 

 inasmuch as they all bear typical stem-leaves, but scattered, not 

 forming a rosette as in the monopodial species. None ot our 

 specimens showed any rhizome, but they were all developed from 

 stolon?, which had rooted and become separated from the mother 



^ Memoirs of the Torrey Botanical Club. V^ol. 2, No. 3, p. 66. 



