/ 



122 



whitish petals up to 2 mm. in length with much narrowed base, 

 were inserted alternately with the four glandular structures 

 which Meehan (Proc. Phila. Acad. Nat. Sci. 51: 95. 1899) con- 

 sidered, probably incorrectly, as representing modified petals. 

 Palmer's collection has plump rounded seeds and should per- 

 haps be identified with L. natans Ell. 



The name Isnardia ascendens Hall in Eaton, Man. N. 

 Amer. Bot. ed. 7: 353. 1836, is of importance, especially since 

 the species came from Albany. It was described as having a 

 "stem obliquely ascending, rigid, (never creeping or prostrate) 

 somewhat angular, reddish, grows with the palustris, but is 

 generally smaller and quite different in appearance." This 

 description seems to have had its origin from Wright and Hall 

 (Plants Vic. Troy, 22. 1836) where Isnardia palustris has the 

 notation, "A variety of this or another species, with an ascend- 

 ing, rigid stem, flowering in August, grows 3 miles west of 

 Albany." This last citation might well place Isnardia ascendens 

 at the site of the station for A mmannia humilis Michx. [Rotala 

 ramosior (L.) Koehne] given by Paine (Cat. PI. Oneida Co. 134. 

 1865) : "Around sandy pools and banks of rivulets in the Pine 

 barrens between Albany and Schenectady, near the Central 

 railroad and Centre station. Rare. July, August." Ludvigia pa- 

 lustris and Rotala ramosior often have a strong superficial re- 

 semblance and the description of Rotala in Britton & Brown 

 (Illus. Fl. ed. 2, 2: 579. 1913), "4-angled stems . . . ascending or 

 erect" corresponds in some respects to the stated characteristics 

 of Isnardia ascendens. Until Hall's specimen can be located, no 

 definite judgment can be passed. Dr. H. D. House, in response 

 to my inquiry, has written that the type of /. ascendens has not 

 been located, and that the fate of Hall's herbarium is unknown. 

 In whatever way the name Isnardia ascendens may be dis- 

 posed of, it is apparent that it does not apply to the estuarine 

 variety of Ludvigia palustris. 



A collection issued by me as Rotala ramosior from Hollow 

 Rock, Tennessee (Svenson no. 4386) differs from all other speci- 

 mens I have seen, in possessing a remarkable woody rootstock 

 3 mm. thick. In this specimen the white seeds average 0.65 

 mm., but the small leaves and arcuate branches place it closer 

 to the slender southern phase of L. palustris (L. nitida Michx. 



