88 J. O. HAGSTRÖM, CRITICAL RESEARCHES ON THE POTAMOGETOXS. 



nearly terete. Ramification from the primary involucral leaves now very inconsi- 

 derable, now exceedingly abimdant with branches of a high degree. The leaves often 

 taper nearly stalk-like at the very base. Apex always provided with a sharp or blunt 

 point, sometimes, it is true, very little obvious but in other instances very prominent, 

 in general abruptly cut out, seldom gradually tapering. Compared with the Oxyphylli 

 and others they are sparingl}^ endowed with sclerenchymatous tissue, though among 

 themselves rather different. Typically the leaves are 3-nerved but in broader leaves 

 a pair of faint secondary nerves appear at the side of the chief lateral nerves 

 especially towards the base, by which such leaves become 5 — 7-nerved. The group 

 is, besides, characterized by the topmost involucral leaves of several species having a 

 tendency towards a structure like that of a floating leaf, by which peculiarity the 

 group shows relationship to the Javanici. The lacunar system along the midrib in 

 a lot of species is more, in others, again, less developed. The lacunar and the mecha- 

 nical systems are in the following crosscut-figures displayed. Oil-cells or vittae at 

 the leaf-base common. 



The stipular sheath behaves differently in this group. In some species we have 

 not the usual, open, convolute ligule, but a connate, closed ocrea, which differs from 

 the ocrea in the sectio I. Connati by being non-plicate aback. By increasing thick- 

 ness of the parts it encloses, it therefore breaks, which takes place in the back, 

 where the tissue usually is thinnest, but in P. mucronatus it breaks both behind and 

 in the front, thus in two lateral parts with strong nerves in the angles. These edge- 

 nerves are three in most species of this group, otherwise one or two, but the ridges 

 are always low and sometimes scarcely perceptible. Beside the lateral vascular bundles 

 the Iigules are strengthened by small mechanic strands both in the front-field (be- 

 tween the two ridges or edge-nerves) and back-field (in case of closed Iigules) or wing- 

 fields (if open), see the fig. 27. After performed office the Iigules deeav by the 

 membranous top, which is lacking a börder-strand, being dissolved and we very 

 soon see a fringe above consisting of the remaining bast-cells. In full-grown, unhurt 

 state the form of the top is rounded, only in P. Hillii rather subacute, though always 

 blunt even here (Fig. 44 E). 



The number of the front-strands is different in the different species, in /'. pu- 

 pillus for instance 8 — 9, in P. Aschersonii, of S. America, there are 3 — 5 only, and 

 so on. 



It is clcar that the condition of the Iigules remarkcd above has a considerable 

 Bystematic value and will facilitate the discrimination of otherwise very similar species, 

 for instance /'. pusilluå L. and P. panormitanus Brv., the former with open, the Latter 

 with connate Iigules, and so on. Of P. trichoides Ch. and Schl., see above! 



The tnrios partly consist of transformed tops of primary branches partly in 

 wholly transformed branches or rudiments of branches; very rarely (hov appear in 

 the rhizomatic paris, i hese being in all species belonging hereto very slender and 

 weak, when present. In most cases true rhizomes are absent. 



The pollen is always spheiie and small-grained. Styles verv slutit, otten with a 

 beak-formed exerescence above, not covered by the low rounded stigma. The sides of 





