54 J. O. HAGSTKOM, CRITICAL RESEARCHES ON THE POTAMOGBTONS. 



ficant connective link between the Coleogetons and EuiDotamogetons. Yet it decid- 

 edly belongs to the latter and can be considered as nearly allied to P. crispiis and 

 zosterifolius. It shows the near relationship existing between the two maingroups of 

 the genus and proves by this, that the primary types of the genus have been sub- 

 merged forms with sheathing leaves. P. densus, another single, remaining species of 

 an isolated section, is a similar evidence as to the ligule, originally being divided 

 into two lateral stipules. 



P. Robbiiisii Oakes. 



In Hovey's Magazine, 7, 1841, 180. — Fig. 19, 20. 



This species rarely fruits and the fruit ripens late, in September — October 

 according to Rev. E. Hill, F1. of the White Lake Region (Bot. Gaz. 1900, 429). 

 However, it can easily be distinguished from any other species by the vegetative 

 organs. The stem is recorded (Hill, 1. c.) to reach a length up to six feet. It is 

 compressed and bicanaliculate as in P. crispus. The furrows especially appear in the 

 peduncle. The stem is branched and elongates from the floral leaves by branches 

 of 1st— 3rd rank. Spiciferous branches are often to be found in the axils next to 

 the primary spike in the way figured as the type C, fig. 2, the plant thus being 

 very rich in spikes. The nervature of the leaves reminds much of that of P. zoste- 

 rifolius. The rounded leaf-base is also mutual. The involucral leaves, again, remind 

 of some Pusilloids by a sometimes occurring slight differentiation between blade and 

 petiole. The perianths are almost circular, concave with short petioles. Style short 

 with a rounded stigma. Fruit more conspicuously petioled than in other species and 

 sharply keeled. 



The pollen of the species exhibits a very strange character in behaving as that 

 of a bastard. The grains, namely, are rather unequal as to their size, commonly 

 of middle size, and mostly sterile. However, it cannot be a hybrid between any 

 modern species. Might not the sterile pollen and the scanty fruiting possibly sug- 

 gest a surviver from times climatically more favorable to it than the present 

 one? Might we not here have before us a dying species, now-a-days maintaining 

 itself principally in a vegetative M^ay? It is likely to be so as regards its area of 

 north-eastern distribution, where it is an immigrant from the time after the glacial 

 period. Its original habitation is the western states of N. America and there we shall 

 certainly find it bears freely fruit. But its distribution and behaviour there is very 

 little known. American botanists should elucidate this question, also of importance 

 geologically. 



Anatomy. The anatomical facts show that P. Robbinsii, like the following, is 

 a decidedly sclerenchymatous plant. — The epidermis of the stem consists of nar- 

 row cells, 4 — 6 times as long as they are broad. Inside it we observe just as in the 

 following groups a more or less complete circle of mechanical strands and, besides, 

 interlacunarly a few cauline strands on each side of the stele (fig. 19, B, cs). This 

 exhibits three well separated portions, the two lateral ones formed by 2 — 3 ± fused 



