154 The Ottawa Naturalist. | December 



While thus cleistogamic flowers appear to be comn-on to a 

 number of species of Viola, there are certain North American 

 species in which they are absent, for instance, V. pedata and 

 rostrata. In V. striata. Canadensis, pubescens and glabella we 

 have observed a few instances where the last developed flower 

 was merely rudimentary, but with no si^ns of producing: any 

 seeds, while such were produced by all the perfect ones on the 

 same stem. And in V. sarmentosa we have not succeeded in 

 detecting a single cleistogamic flower in our herbarium specimens, 

 and not in V. Langsdarfii either. We hope, however, that Cana- 

 dian botanists will re-examine these species in the field, more 

 especially the two last mentioned, since it is very important to 

 learn something about the structure of the cleistogamic flowers in 

 general in others than those described above 



VARIATION IN LEAF-OUTLINE. 



When numerous leaves develop from the same bud as in the 

 monopodial, acaulescent violets, certain variation becomes always 

 more or less noticeable in the leaves. Those that develop first, 

 before the flowers, are frequently different from the later ones, 

 and in certain sptcies, V. papilionacea and pahnata, for instance, 

 the first of these are generally cordate or reniform, but entire, 

 while the later ones, sometimes, are more or less deeply lobed. 

 In V. sagittata the variation in leaf-outline is quite considerable, 

 and we have, sometimes, noticed a number of forms upon the 

 same individual during one season, from the oblong-ovate to the 

 lanceolate, with ihe base hastate, or from the deltoid, entire to 

 the deeply lobed, the latter being characteristic oi the so-called 

 V. emarginata. Such variation seems largely due to the position 

 of the leaves in the bud, but there are, also, cases where the 

 nature of the surroundings seems to affect the leaf-shape. Viola 

 emarginata. for instance, does not develop the deeply cut leaves 

 except when it grows in rich soil and in shade ; in open places 

 and in sandy soil the leaves become entire and often quite narrow 

 like those of V. sagittata. V. palmata has always the later leaves 

 deeply lobed, when growing in woods, while V. papilionacea 

 shows a pronounced lobation, when observed in damp, shaded 

 places, along creeks, etc. 



