182 THE JOUKNAL OF INDIAN BOTANY.
Numerous gatherings ; all c. fr.
I have examined the types of Hooker's Gymnostomum involutum and G. cyliridricum, and I am unable to find any distinction. Hooker possibly was misled by an error he had made in the figure of G. involu- tum, where he depicted the leaf as entire whereas in the type specimen it is actually denticulate, as in G. cylindricum. The degree of involu- tion of the leaf margin when the leaves are in the moist condition varies very much in different specimens, and probably according to the age of the leaf, and too much reliance must not be placed on this as a separating character.
After examining a large number of specimens that would come under H. tenocarpa Een. and Card. I have finally come to the conclusion that it cannot be retained, at any rate as a species. It was founded, somewhat tentatively, as distinct from H. cylindrica on the strength of the taller stems, and the leaves scarcely involute when dry; the narrower, longer capsule, with a finer and longer beak to the lid. All these characters, however, I find very variable ; this applies very markedly to the form of the lid, which in the type of Hooker's G. in- volutum (Nepal, Wallich) shows lids from short and obtuse to almost twice the length and finely subulate ; and in the South Indian plant the form of capsule and lid are so variable that I can draw no line of distinction between H. stenocarpa and H. involuta. The plants vary very much also in size, and length of stem.
In all probability further reductions will have to be made in the genus.
Hyophila Walked Broth. Stones, Karwar (6366) c. fr.
The nerve is described as percurrent or very shortly excurrent while in Sedgwick's plant it rarely reaches quite to the apex. But I find this to be equally the case, frequently at least in original speci- mens of H. ivakeri at Kew.
Barbula consanguinea (Thw. and Mitt) Jaeg., forma.
Bole of coconut palm, spice gardens, Shiggaon (3487).
Although a tall, sterile plant, with the stems 2 cm. high, and therefore differing considerably in habit from the normal plant, the structural differences are very slight (leaves a little narrower below, wider and less acute at apex) and I think it is certainly a form of B. consanguinea.
Barbula dharwarensis Dixon sp. nov. (fig. 7).
Sordide virens, caespitosa, terrestris ; caulis circa 1cm. altus ; folia ebasi paullo dilatata ling ulato-lanceolata, apice piano, obtuso, vel sub- acido apiculato, mar gine alter o inferne leniter recurvato, ceterum piano, integer o ; patula, sicca crispato-incurva, 1 — 1.5 mm. longa. Costa
