ON ANOMOCLADA. 195 



A few salient and easily-observed characters suffice to distinguish 

 the Uyo In 0. Sphagni the trailing stems grow out at the apex 

 almost indefinitely — I have seen them six to eight inches long — 

 lengthening by repeated archings, and rooting by ilagella at the de- 

 scending flexures, and are clad throughout almost uniformly with 

 leaves of equal size ; the few branches preserve the same habit and 

 character, nor are they ever seen erect and gemmiparous. In 0. 

 deniidatiim we have first an intricately-branched leafless caudex, or 

 rhizome, which puts forth assurgent arcuate branched stems that 

 scarcely ever root at the decurved apex ; and both stems and branches 

 are (normally), short, linear-lanceolate in outline, from the leaves 

 beiui^ longest at the middle of the branch and decreasing in size 

 towards either extremity ; but there are nearly always present a few 

 erect branches, gemmiparous at the apex, on which the leaves decrease 

 in size from the base upwards, while the associated folioles increase, 

 80 that the upper leaves and folioles are nearly equal in size ; whereas 

 the folioles that are scarcely ever absent from the ordinary branches 

 and the stem are many times smaller than the adjacent leaves. Eut 

 on 0. Sphagni there are either no folioles at all, or only a few minute 

 ones near the apex of the ^tem and branches. The leaves scarcely 

 differ in form in the two species, but are rather smaller and more 

 pellucid in 0. denudaium, and never secund (as often in 0. Sphagni). 

 Nor is there any marked difference in the involucres and perianths, 

 as the above descriptions will show ; but, on the whole, there is no 

 reasonable doubt that the two species should be held distinct.* 



In 1845-6 I gathered Odontoscliisma denudaium abundantly, with 

 perianths and male flowers, along the western foot of the Pyrenees, 

 especially at the entrance of the Vallee d'Ossau, where it grew on the 

 decaying heads of polled Chestnut-trees, and on rotting stumps of other 

 trees, along with (but rarely upon) Leiicohryiim glaucum. In England 

 I have seen it only on crumbling sand-rock near Tunbridge Wells ; 

 and in the Andes — -only once, and in small quantity — on moist quartz- 

 ose sand, on Mount Campana, in Peru, at about 1400 metres eleva- 

 tion. So that I have seen it, for the most part, only on dead vegetable 

 or mineral matters ; while O. Sphagni prefers to grow on living 

 plants — in European turf-bogs on Sphagna and Leucohryum ; in 

 Amazon forests at the foot and on the roots of trees, or creeping over 

 Mosses on prostrate trunks. 0. denudaium is given in SuUivant's 



* Lindberg finds the leaf-cells of 0. denudation verruculose, those of 0. 

 Sphngni smooth ; but I find a slight and variable degree of verrucosity in the 

 cells of both. This appearance depends on the presence of minute pustules on 

 the outer surface of the cells, mostly equidistant and sometimes closely set 

 in lines (striote). They may be beautifully seen in Trichocolea, in some Micro- 

 pterygia, &c., and are probably never quite absent from a,nj Jungermannidea. I 

 venture with great difiidence topredict that the cell-pustules will one day be proved 

 perforated at the apex, and that it is through these perforations the cells imbibe 

 fluids so readily. I have watched, under a high power, the behaviour of a 

 dried leaf on the application of water, and noticed that the pustules of the cells 

 seemed to first take it up, also that (apparently) they previously emitted minute 

 air-bubbles. Perhaps someone with younger eyes and better instruments may 

 be induced to repeat my observations. The term endosmose is one of those con- 

 trived to conceal our ignorance of Nature's actual processes ; for a membrane per- 

 meable to water is plainly not absolutely imperforate. 



2 



