KUNGL. SV. VET. AKADEMIENS HANDLINGAR. BAND 55- N:0 5- 61 



North as to Yeniseisk 58° 20' N. Lat. where it is gathered by H. W. Arnell in 1876. 

 (see Scheutz, Pl. vase. jenisei. p. 163! hb. Stockholm.) with a year isotherm of about 

 — 4° C. Such a curve towards the north in these regions may belong to several 

 other species of about the same northern limit, undoubtedly dependent on the com- 

 paratively high summer temperature (July isotherm + 20° C). From this northern 

 limit it seems to spread in Europé and A sia, through Africa : Egypt, Syene, Kotschy 

 (hb. Stockholm, et Uppsal.), var. obtusijolius f. angustifolius Fieb. Natal, in rivulo 

 .Umschlangwe pr. Phoenix, 93, R. Schlechter N:o 3122 (hb. Ziirich. et Stockholm.) 

 (see also Ar. Bennett in Flora trop. Afr. VIII: 2, 1901!), and Australia. The distri- 

 bution in N. America is little known. Arizona is mentioned; also Canada. In our 

 Museums specimens are found from Delaware, Commons (hb. Stockholm.) and Massa- 

 chusetts, Arlington, coll. Morong (hb. Stockholm.). From Central America and 

 South America it is not recorded. 



P. crispus L. x perfoliatus L. 



Fryer, On a nevv Brit. Potamog., in The Journ. of Bot. 1891, 289 tab. 313. - 

 P. undulatus v. Cooperi Fryer ; P. Coopéri Fryer, 1. c. 290. — P. cy matodes Asch. 

 et Graebn., Synops. mitt. Fl. I, 1897, 337. — P. cymbijolius Fischer, Beitrag zur 

 Kenntn. bayer. Pot. in Mitteil. n:o 31 d. bayer. Bot. Ges. 1904, 366. 



Of this hybrid I have seen four forms which can be classified into two series, 

 one with the leaf serrature crispus-like with real incisions in the margin (Fig. 21, L), 

 the other jierfoliatus-like with one-celled denticles on the margin (Fig. 117). The 

 Ascherson— Graebner arrangement is more vague and difficult to understand and 

 apply. 



Stem always compressed and more or less double-furrowed as all the crispus- 

 bastards. Leaves always provided with teeth or denticles, 7 — many-nerved, 

 broad or narrower, short or more elongate; leaf-apex obtuse and always more or 

 less gutter-shaped and perfoliatus-Wke. Nerve-spaces either short as in P. crispus, or 

 somewhat longer as in the other paren t-species. Ligules always very deciduous. 



The stem-anatomy has three singularities: 1) Endodermis always of O-cells; 2) 

 Cortical bastbundles either entirely wanting or seldom represented by one or two 

 interlacunar strands with a disposition for running subepidermally (an inheritance 

 from P. perfol.); 3) Along the epidermis there never appears more than a single i. e. 

 one-celled layer of bark-cells. By these anatomical facts this hybrid can easily be 

 distinguished from the very similar hybrid P. undulatus, where the stem is always 

 much influenced by P. prcelongus, and therefore has: 1) an endodermis of more or 

 less conspicuous £7-cells, 2) the subepidermal bark-layer two-celled at least in several 

 places, and 3) vascular and bast-bundles in the bark at least in one well developed 

 interlacunar cirele, but on the other hand scarcely any subepidermal strand, thus 

 an arrangement in opposite direction to the foregoing hybrid or from epidermis to- 

 wards the central axis. 



