KUNGL. SV. VET. AKADEMIBNS HANDLINGAR. BAND 55. N:0 5. 



:o7 



The transformation, however, is not very significant. The internodes cease to grow, 

 increase somewhat in thickness and form a reservoir for starch. The leaves grow a 

 little shorter, having for the rest the usual shape found in young branches. 



This plant resembles at a first glance our common pusillus, or still more the 

 main-form of P. panormitanus but differs on closer examination from both of them. 

 The stem is merely about 20 cm high, sending forth now and then from the base 

 a single fruiting branch according to the ^-type, otherwise only endowed with 

 buds or undeveloped branchlets, or branchlets transformed into 

 turios. Nor do we find any propensity in this plant to prolong 

 itself by spike-bearing branches from the involucral leaves. I never 

 saw more than one very short and insignificant spiciferous branch. 

 Anatomically the stem is constructed as P. fusillus. In the peduncle, 

 being a little thicker than the stem, the four vasal bundles run 

 together, but beside the pseudo-hypoderma there are also scattered 

 subepidermal bast-bundles. The leaves are somewhat broader below 

 the middle, at the most 1,75 — 2 mm, toward the top 1 mm. At 

 this enlargement of the blade there are often 5 nerves, but usually 

 the leaves are three-nerved. The nerves join the mid vein at about 

 a leaf-width's distance from the very apex. The midrib is strong and 

 generally accompanied by two rows of lacunae almost up to the very 

 apex. The apex is obtuse with only a little wart, but on the young 

 branches described above the leaf-apex is more keenly cuspidate, 

 by which fact the leaves can be said to be dimorphous. The front- 

 field of the ligules is 7— 9-fibrous. The perianth leaves have the 

 usual rounded form. The fruit is both in size and form different 

 from the pusillus-ivait, but on the other hand very like the fruit of 

 P. 'panormitamis, with which P. exiguus, however, is not to be con- 

 founded for the sake of the obtuse stem-leaves, the turios, the sub- 

 epidermal bast-bundles in the peduncle etc. It shows also great rela- 

 tionship to P. antaicus by the ligules, the obtuse leaves, and the 

 course of the nerves, but the cross-veins are in the new species 

 joined in almost right angles to the longitudinal nerves, the mode 

 of branching and prolongation is another etc. 



Distribution. Africa, Cape Town, 53, coll. N. J. Andersson 

 (hb. Stockholm.). 



e.viguiis 



Fig. 42. 

 Hagstii. a, Transverse 

 section of a stem- leaf at 

 the middle, sir, strands, 

 n' lateral nerve, \°. B, 

 Top of a stem-leaf, ^-^ . 

 C, Top of a branch-leaf, 

 y. D, Fruit, outlined, 

 lateral view, y. 



P. limosellifolius Maximowicz. 



In KoRSHiNSKY, Plantse Amur. etc. in Act. Hort. Petropol., 1893, 393. — 

 Fig. 43. 



This species also joins the Pusilli connati by all its chief properties, but it 

 cannot yet be said to be fully known. It is not present in the Swedish Museums. 

 The specimens I here refer to are preserved in the Museum of S:t Petersburg and 



