KUNGL. SV. VET. AKADEMIENS HANDLINGAR. BAND 55. N.O 5. 



109 



P. foliosus, as well as by the peduncle being very short. Ooramon to these two 

 species are also many other characters, for instance the few-flowered spikes and the 

 crest-formed outgrowth in the dorsal and ventral sides, which formations seem to be 

 more conspicuous in the form without oil-cells. In size, again, the fruit reminds of 

 P. zosterifolius. 



The leaves are endowed with a well developed lacunar and mechanical system 

 according to the figure below. Ligules more tapering than is usual, their front-field 

 10— 13-fibrous. 



Distribution. N. America, Manistee, Mich., 82, T. Morong (hb. Stockholm.). 

 The supposed Asiatic occurrence of this species is most doubtful. 



str 



rmTA é ± 



P. Gayii Ar. Bennett. 



Bemerkungen iiber die Arten der 

 Gått. Pot. etc. in Ann. naturhist. Hof- 

 mus. Wien, 1892, 293. — Fig. 45. 



I have seen a great deal of indi- 

 viduals of this species, but none of them 

 with so broad leaves as is stated in the 

 original description, 4 — 9 mm. The leaf- 

 width is about 3 mm and this is also 

 the fact concerning specimens deter- 

 mined by Bennett himself. Several 

 other statements in the description (for instance of the affinity etc.) compared with 

 material deposited in Sweden seem to me very stränge. 



Although developing plenty of spikes the species has not been met with fruit- 

 ing. This makes me think that the plant may be of a hybrid origin, an opinion 

 which is supported by the fact that the pollen to the greatest part consists of 

 wrinkled, angular, and empty grains. The stigmas on the contrary, that I have 

 examined seem in no way to be abortive or deformed. We still know the South 

 American species of this group quite too little to venture a proposition in any direction. 



Characteristic to the species is, that it willingly elongates with spike-bearing 

 branches above, yct not dichotomously, like for instance P. obtusifolius, but so that 

 the branchlet of the lower involucral leaf is generally left undeveloped while the upper, 

 stronger branch takes the charge of the stem-prolongation and development of 

 spikes. By this there exist a great number of spikes (I have observed as many as 

 five) ranked pseudo-laterally. No tendency to produce semi-coriacious leaves can be 

 traced in this species. Turios of transformed short branches are found at the base 

 of the stem. Their lowest ordinary leaves narrow, but otherwise of usual shape. 

 The next following 2—3 leaves short and narrower still, the innermost ones l,i mm 

 broad, a little surpassing their sheaths and somewhat obtuser than the inferior leaves, 

 the whole turio about 25—30 mm in length. 



Fig. 44. I'. Hillii Mor. A, stem-leaf, base (a) and top (/;), * . 

 B, Transverse scction of the lacunar part of a stem-leaf, vascular 

 (mi) and bast bundles (str) hatehed, about S j°. C, Transverse sec- 

 tion of the peduncle showing the site of the bundles, y. D, Pistil, 

 a, side-view, b, Stigma from above, f. E, Ligule, J. 



