1903.] NATURAL SCIENCES OP PHILADELPHIA. 677 



Similar leaves 50x60 mm.; petioles 100-110 mm.; scapes 90-100 

 mm. ; cleistogenes horizontal or slightly ascending, 20-40 mm. long. 



Fruiting plants, Jmie 17, 70x80-80x100 mm.; petioles 225 mm. 

 long, fruiting cleistogenes 50-60 mm. long. 



A glance at PI. XXXIII will show to what extent the leaves of this 

 violet vary; all sorts of forms may occm- on one plant, or we may have 

 plants with the leaves all uncut. 



The latter are supposed to be the basis of Willdenow's Viola sororia. 

 I fail to see, however, on what grounds this can be regarded as a separa- 

 ble form, even as a subspecies, as is done in Britton's Manual. On 

 the other hand, if we regard it as the same as dilatata, it must supersede 

 it, being an older name. To my mind the same arguments apply here 

 as in the case of V. obliqua Hill. Nobody can ascertain what the 

 original plant was, and in view of the various uses to which the name 

 has been put, it had best be discarded as unrecognizable. 



The several other varieties given by Elliot are probably not separable 

 from this, and Schweinitz specimens which I have examined prove 

 his triloba to be the same thing, while hderophylla Muhl. and asarifolia 

 Pursh, while probable referable to this, are neither of them availal^le. 

 This is one of the most common woodland violets found about Phila- 

 delphia, and may be recognized from any (except the next) by its 

 uniform pubescence, which is especially marked on very young leaves, 

 though disappearing to some extent on the large foliage of late 

 summer. 



16. Viola palmata variabilis (Greene). 



Viola variabilis Greene, 1902, Pittonia, V, p. 91. 



Range. — Apparently higher ground, nearer to the mountains or foot- 

 hills. 



Habitat. — Dry woodlands. 



Description. — A close ally of the preceding, but smaller and much 

 more pubescent, quite villose on young leaves. Flowers similar but 

 leaves more deeply divided and with fewer uncut leaves. Specimens 

 from Argus, Bucks county, Pennsjdvania, May 8, 1903, Steward- 

 son Brown, No. 5,110, and June 8, Dr. C. D. Fretz, No. 5,111, Herb. 

 W. S. 



These plants have been identified as V. variabilis by Prof. Greene, 

 while the former, whicli I have called dilatata, he considers not suffi- 

 ciently pubescent and in other ways unlike this form. It may be that 

 further study with more material may show that they are not dis- 

 tinguishable. 



