pinnate, sessile, free, alternate, patent, 1 line or a little more long, sub-quadrate with a single deep notch at apex and nearer to the inferior side, slightly arcuated on the superior side, and very finely and closely toothed on its outer corner and round it a little way on the apex: sac, or torus, sub-terminal on both main and lateral branchlets, sub-globose or broadly oval, 1½–2 lines long, densely hirsute-hispid, colour light brown.

Hab.—On shaded clay banks and on rotten logs near watercourses in thick wood near head of the River Manawatu, North Island; 1879–1881.

A species possessing close affinity with Gymnanthe tenella, Taylor, and Marsupidium knightii, Mitten.

This species I have long known in its barren state; and although it appeared to be very nearly allied to Gymnanthe tenella, Taylor, of New Zealand and Tasmania (vide “Fl. Tasmaniæ”), yet I could never quite believe it to be the same; and now that I have found it pretty copiously in fruit, I am certain of its specific distinction. G. tenella is fully described by Taylor (who established the genus on that species), in “Lond. Journal of Botany,” vol. iii., p. 377 (and in “Syn. Hepatic.,” p. 192), and a drawing of it is also given in the “Fl. Tasmaniæ.” In foliage and in size and in manner of growth the two plants are very much alike; still, the leaves of this species are not so closely set, and have many more and finer serratures at the apex (9–10) than there are in that one, which usually bears but three. But the chief distinction is in its sac or torus, which in G. tenella is described as “elongato obconico striato”; while in this species the same part is densely shaggy, almost echinate when fresh.

In the “Handbook of the New Zealand Flora,” p. 520, G. tenella, G. saccata, and G. urvilleana, with other Hepaticæ, were all lumped together under the one species—G. saccata. (This, to me, who had formerly collected them all in New Zealand, seemed surprising, as I could not discern much of a close resemblance between them.) Subsequently, however, Mitten broke up the genus (though but a small one) into several new genera,^* and in so doing not only restored the three above-mentioned species of Gymnanthe (which I was pleased to see) but even separated them into distinct genera.

It is not, however, stated in which of those new genera G. tenella is now placed; possibly in Tylimanthus; but this plant of mine will, I think, be found to rank naturally with Marsupidium, and seems pretty closely allied (judging from the short description) to Mitten's new species, M. knightii (p. 753, l.c.), which is also a New Zealand species.