76 J. O. HAGSTROM, CRITICAL RESEARCIIES ON THE POTAMOGETONS. 



is the case with several other nearly allied species. One or a couple of subepi- 

 dermal strands can also be perceived in the peduncle. — E piller mis-cells alvvays 

 rather short, usually 2—4 times as long as they are broad. 



The leaves are very narrow (£— | mm) always three-nerved and tapered at 

 the base. Apex acute except in the involucral leaves, of which at least the topmost 

 one usually has a more or less obtuse apex. The lacunar middle portion of the leaf 

 is abundantly endowed with sclerenchyma. Its anatomy as seen in the fig. 29, C. 

 The inner leaves of the iurios hidden by the sheaths are verv short, obtuse or round- 

 ish-obtuse, pointless, 7 — 8 in number, the outer or lower ones, 2 — 4 in number, are 

 free, a little protruding, and in the basal part semiterete, by which they differ from 

 the other branch-leaves. 



The ramification of this species is usually very abundant. According to J. Gay 

 very short rudimentary branches grow out from the axils, on which låter on the 

 longer branches are formed in an unlimited number as branches of the second rank 

 while the rudimentary branch grows on. Those longer branches, however, are said 

 to be found only on the side of the short branch turning to the stem, in the axils 

 of the 2nd, 4th, 6th etc. leaf, whereas the 1st, 3rd, 5th etc. leaf are said to 

 be barren. This explanation given in 1854 (see B ull. Soc. Bot. France of that year, 

 p. 46 et seq. !) has as late as in 1907 been adopted by P. Graebner. I cannot 

 join this opinion. The mode of ramification of this species is fully concording with 

 that of P. vaginatus (vide fig. 11, D, pag. 32!) that is to say the branches are of 

 the first, second, third etc. rank in relation to the main-axis, see also Raunkijer, 

 Danske Blomsterpl. 1896, p. 37—38! 



Another character combining this species with the Co?eo(7C/on-species is that 

 the strongest of the buds of the two involucral leaves is the lower one. Those buds 

 do not, however, come to any rich evolution as for instance in /'. pusillus, but the 

 spike-production is generally accomplished with the primary spike. Nevertheless the 

 species becomes very rich in spikes since the numerous branches beneath the primary 

 spike are in general spiciferous according to the type C, Fig. 2. This peculiarity 

 gives to the species a characteristic appearance among the other narrow-leaved spe- 

 cies, corresponding to the state or aspect of P. vaginatus withiu the subgenus 

 Coleogelon. 



The species does not vary considerabl}'. Some authors have distinguished 

 varieties with three-nerved leaves, but the leaves are on doser examination always 

 three-nerved. Sucli varieties as Trimmeri Casp. and capillaris Fisohbr, recorded as 

 three-nerved and by this fact separated from the main-form, are probably bastards. 

 The species has, namely, a formerly not sufficiently observed tendenoy of crossing 

 with a great many species, by which fact tnany aberranl forms get a Batisfaotory 

 explanation. 



The fruit, however, varies as to the development of bossea or protuberanoes 

 the producing of which P. irich. is inueli [nolined to. This peoularity has, however, 

 not any great systematie importanee, as the same individual otten shows fruits with 

 an appearance different in this respect. Chamisso'8 desoription evidently refers to 



