234 J. O. HAGSTRÖM, CRITICAL RESEARCHES ON THE POTAMOGETONS. 



which a transitory form to the semi-terete submersed leaf is formed. This reduction 

 mostly takes place when the plant grows in tufts and many shoots spring from 

 the same rootstock; but is also met with in solitary specimens standing alone, and 

 not only in greater lakes with vehement billows but even in smaller ones. For the 

 purpose of assimilation short branches are developed with always ordinär j r membranous 

 loaves. This peculiarity is characteristic for the species, and therefore we always 

 find one or a few semi-terete stem-leaves, at least at the base. If this reduction 

 is continued higher up to the stem's middle the leaves or phyllopodia grow verv 

 long; at the top, again, they have the usual length anew. This reduction of the 

 laminoe cannot be regarded as a variety in usual sense as the same rhizome obviously 

 is able to produce shoots of much reduced form beside such ones bearing leaves of 

 usual width. The reduction of lamina passes on in about the same proportion from 

 both ends or in a little higher degree from the apex. When Z«ce?is-leaves, consequently, 

 on account of this reduction grow longpetioled, they also have a corresponding 

 elongated cusp. Among other differences from a pure P. lucens the large-leaved 

 forms of the hybrid P. Zizii exhibit is the property that, with more or less long 

 upper leaf-stalks, they want the prolonged cusp peculiar to the lucens-le&i (influence 

 from P. gramineus). 



Leaves in general ]l-nerved in such way that three on either side are running 

 near the margin and two at a greater distance from one another and the midnerve 

 through the middle part. The nerve-spaces formed by the crossveins are, thus, in 

 this species typically short and broad in the longitudinal direction of the leaf, in P. 

 gramineus, again, more extended, often longer than broad. This peculiarity is also 

 traced in the hybrids lucens X natans and gramineus X natans in the leaves constructed 

 as submersed leaves. Besides the leaves are always petioled, branch-leaves always 

 very shortly, stem-leaves usually some mm longer petioled or considerably longer 

 if the lamime are reduced. On the margins the leaves are furnished with close, small 

 sinuses, each with a stout, acute, forward directed topcell, by which a beautiful 

 denticulation is formed. This denticulation is always suppressed in crosses with 

 P. natans, but is maintained, though much debilitated by hybridizing with P. yrwlongus, 

 which shows that it properly belongs to the thin, membranous, submersed leaves, but 

 is extraneous to the semi-terete ones. On the stipules etc, pollen, pistil, and fruit, 

 sec O. Hagström in Neuman. 1. c. 1901, 797! Endocarp not keeled. 



Concerning the stem-anatomy you may compare C. Sauvaok \r, Notes biol. in 

 Journal de bot., 1894, n:o 6. The deviation mentioned there on p. 119 with regard 

 to the incrassation of the endodermis-cells, o-cells in one internode while all the 

 otliers liad w-cells, must not be conccived as if this species could have now u-oells 

 now o-cells in the stem-endodermis. The deviation is only an accidental return to 

 t lic primitive type, which obviously haa becn o-cells. In the very numcrous specimens 

 1 have examined I never found this otlierwisc verv interesting fact. The condition 

 ia quite the same with the behaviour of the vaseular bundles of the stelc. Typically 

 bhese bundlee are arranged so that the xylem ])arts of the median bundles unite 

 and by degrees diasolve into a oommon xylem canal. Sometimes, however, the fusion 



