380 Transnctloiis. — Botany. 



Hab. South Island — Lochnavar, Valley of the Poulter, J. D. Enys. 



An extremely slender species, allied to C. nana, Hook. f. That species 

 however differs in the wider branchlets, more compact, stouter habit, 

 racemed flowers and broader pods. 

 Carmichcelia munroi, Hook. f. 



Leaves minute, cuneate or obovate, emarginate, silky, gradually suc- 

 ceeded by minute scales. Pod |"-£" long, slightly falcate, rarely straight, 

 with a short straight or oblique beak, remarkably turgid, valves corrugated. 

 Seeds 4-5 mottled. 



I suspect that the leaves described above are preceded by others of larger 

 size. 



Carmichcelia ivilliamsii, T. Kirk. Plate XXXII. 

 Trans. N.Z. Inst., xii., p. 394. 



A further supply of specimens from Archdeacon W. L. Williams enables 

 me to offer a plate of this fine species, and to add a description of the fruit 

 to the characters already given. None of my pods however are fully mature : 

 the most advanced is a single pod collected by Mr. Buchanan at Hicks Bay, 

 and for which I am indebted to Dr. Hector. 



Pods on stout erect pedicels, slightly turgid, oblong or obliquely oblong, 

 with a long straight beak, l"-lj" including beak. Seeds 9-10 red, mottled 

 with black. 



Carmichcelia kirkii, Hook. f. Plate XXXIII. 

 Ic. PI., t. 1332. 



Branches few, distant, extremely slender, 2-3' long, or more, terete, 

 grooved. Leaves few, pinnately 3-5-foliolate, |"-1" long, leaflets sessile, 

 orbicular, emarginate, I" long, glabrous, racemes 3-5-flowered, lax, rhachis 

 and pedicels slender, bracts ciliated. Flowers J" long, calyx glabrous, teeth 

 acute, vexillum 2-lobed, orbicular, wings longer than the keel, ovary glab- 

 rous. Pod i" long, ellipsoidal, with a long perfectly straight subulate beak, 

 valves faintly reticulate, seeds 2, replum stout, broad. 



Hab. South Island, Cardrona Valley, T. Kirk. Otepopo, D. Petrie. 



A very distinct species, differing from all others in habit, flowers, and 

 fruit. The long flexuous branches are unable to support their own weight 

 and usuaUy become interlaced with adjacent shrubs. 

 Carmichcelia pilosa, Col. 



The branches of the North Island plant are much compressed and 

 usually distichous, but on the Canterbury plains they are frequently plano- 

 convex, or even terete and fastigiate. I can, however, detect no difference 

 in the pods ; the ovary varies in the degree of silkiness ; the same branch 

 may produce specimens with but few hairs, while others may be villous. 

 In most cases the hairs speedily fall away after fertilization has taken 



