190 Rhodora [August 



and highly variable American plant which is here called P. gramineus, 

 var. graminifolius. In Freyer'a Potamogetons of the British Isles 

 typical P. gramineus L. ( - P. heterophyllus Schreb.) is figured and 

 described as having the upper stipules strongly divergent and the 

 short peduncles conspicuously thickened at summit, while P. gramini- 

 foltUS is illustrated with more appressed-ascending stipules and 

 elongate barely club-shaped peduncles. All American material in 

 the Gray Herbarium and the herbarium of the New England Botan- 

 ical CMuh, altogether about 300 sheets, agrees with P. graminifolius 

 in these characters, and typical /'. graminetu or P. heterophyllus is 

 rare if not quite unknown in North America. Frcyer indicates 

 differences in the fruit, although it is significant that in his description 

 he was obliged to quote from Morong the supposed distinctive char- 

 acters of the fruit of P. graminifolius. These differences, however, 

 do not appear constant and it is noteworthy that many American 

 plants, otherwise good P. graminifolius as treated by Frcyer, have 

 the fruits quite like his illustrations under P. graminetu. It seems 

 best, therefore, to consider P. graminifolius a strong variety of the 

 complex P. graniinctts, as lias been so generally the practice for a 

 full century by students of the Pondweeds. 



** P. gramineus, var. sFATniLAEFOKMis Robbing in Gray, Man. 

 ed. 5, 4S7 (1867). /'. spatkaeformis Tuckerm. ex Bobbins, I. c. 

 (1867); Bennett, Journ. Hot. xxxviii. 130 (1900); Fernald, Rhodora, 

 viii. 224 (1906); Robinson & Fernald in Gray, Man. ed. 7, 74 (1908). 

 P. tarians Morong ex Frcyer, Journ. Hot. xxv. 308 (1887), xxvii. 

 33, t. 287 (1889); Frcyer, Pot. Brit. Isl. 07, t. 41 (1915). P. spathulae- 

 formis (Robbing) Morong, Mem. Torr. Hot. CI. iii. pt. 2, 20 (1893), 

 but hardly t. 35 (with attenuate and therefore quite uncharacteristic 

 submersed leaves). P. Zizii x gramineus Ascbers. & Graebn. 

 Synop. Mitteleur. Fl. i. 327, in part (1897). P. spathuliformis Asch. 

 & Graebn. in Engl er, Pflanzenr. iv. Fam. 11:91 (1907). — Apparently 

 local, collected only once in DlOBY Co.: brook with muddy bottom, 

 outlet of Midway (Centreville) Lake. 



Var. spathulaeformis was proposed by Robbins with doubt as to 

 its exact affinity because his material from Mystic Pond in Middle- 

 sex Co., Massachusetts, was sterile. Newfoundland material from 

 two stations closely matches the original Mystic Pond collections 

 and the material from Grand Falls, Newfoundland {Fernald k. Wie- 

 gand, no. 4475) might well have formed the basis of the plate of /'. 

 mrians in Freyer's Potamogetons <>f the British Isles. Both New. 



