NOTES ON POXDWEEl>s. 35 



and that from the River Boyne closely agree. Allowing for the 

 evidently different local conditions under which the plants have 

 grown, confirmed "splitter" as I am, I can find no single character 

 by which the three plants may be separated. Hence I consider the 

 Irish " P. lonchitis" to be also a state of the species above described 

 as P. (jramineus L. v. graminifolius Fries. 



This name was arrived at by the examination of a fine series of 

 Scandinavian forms of P. heterophil I us named and arranged by Dr. 

 Tiselius ; on looking them over I found several varietal names 

 added to the general specific one of fruunneus L. (used by many 

 continental botanists as being prior to luterophylliis Schreb.), and 

 amongst them was that of grammiftJitu Fries; then, selecting 

 specimens apparently belonging to this form, without further 

 reference to Dr. Tiselius' labels, I found a very natural group of 

 forms all of which bore a strong resemblance to the newly observed 

 type-plant of Pidley Fen, and all of which, on reference then being 

 made to the labels, bore the varietal name, "yramitiifoUMB" 1 To 

 these the specific likeness to my plant was seen to be almost com- 

 plete, and I had no hesitation in ranging my plants under the same 

 varietal name. It will probably be understood why I use it in this 

 paper, which is written not so much to describe a form which 

 seems little known, or unknown, as a British plant, as to show, in 

 the first place, how extremely wide a state of a Potamogeton may 

 depart from its typical form, and yet suddenly revert to it under 

 slightly altered conditions of growtli ; and also to serve as a sort of 

 " object-lesson " to illustrate the method I would recommend 

 students to adopt in naming local forms tliey may happen to find. 

 Probably they will make mauy mistakes, as probably I have done 

 in the case here related; but they will, by the process recommended, 

 obtain more real knowledge of the genus than they possibly could 

 by sending a fragment of a plant to be compared with another 

 fragment, about the real specific value of which the original namer 

 likely enough knew little or nothing. In well-defined species no 

 method of naming a plant can be safer than matching with a well- 

 authenticated typical specimen ; but it is quite otherwise with any 

 form belonging to a " critical" group in this genus. 



P. gr amine m L. seems to divide itself into two natural sections, 

 gramimfolius Fries and heterophyllus Schreb., between which a com- 

 parison shows the following differences : — The fruit in v. hetero- 

 phyllus seems to be larger, less rounded on the dorsal margin, and 

 less prominently keeled, and the habit of growth is less like that of 

 P. Zizii than in v. gramimfolius. 



But thus far I have only got to a fitting place and name for my 

 new form ; the next question was, — What relation did it bear to 

 certain segregates I had proposed and named in the pages of this 

 Journal ? This, I must confess, I approached with some feelings 

 of dismay. For I had, during the four years of watching and 

 collecting the Pidley Fen plant, noticed several states (as I now 

 know them to be) which respectively resembled, more or less, all 

 the other heterophylliis-forws of the fens, and I had suggested, or had 

 suggested to me, for them such names asfatcatus, varians, fluctuant, 



d2 



