T. Kirk. — On Hymenantliera. 511 



membranous ; the stipules are usually small and fugacious. 

 The flowers are hermaphrodite or dioecious ; solitary, or more 

 usually fasciculate and shortly pedicellate, rarely solitary ; 

 they are rarely produced in the axils of the leaves, most 

 frequently on the naked parts of the branches. The sepals 

 and petals are obtuse or rounded at their apices, or the petals 

 may be narrowed above and shortly revolute. The androe- 

 ciuin consists of five sessile anthers which open longitudinally 

 and form a ring surrounding the style ; their connectives are, 

 however, connate, and are produced into a rounded or sub- 

 acute ciliated membrane above each anther, with a curious 

 obovate dorsal appendage. The ciliated processes and their 

 dorsal appendages vary in shape in the different species, but 

 not sufficiently to afford distinctive characters. Stigmas 2, 

 rarely 4 or 3, divergent, styles very short. The fruit is a 

 spherical 1-celled berry of a deep purple colour, or rarely 

 white. The seeds are 2 in number, rarely 4, 3 or 

 sometimes solitary by abortion ; they may be ovoid, plano- 

 convex, or convex and pointed with one or two angular faces 

 at the base and a strophiole which may be very slightly 

 developed or large and distinctly cupular. The cotyledons are 

 orbicular in all the seeds examined by me. The most im- 

 portant differential characters are those afforded by the seeds 

 taken in conjunction with the leaves. 



Baron von Mueller appears to have been the first to draw 

 attention to the plano-convex form of fruit (Plant. Vict., i., 69), 

 but did not attach to it the importance which it seems to me 

 to merit. 



It will be seen that H. dentata, E. Br., is only represented 

 by its variety angustifolia ; I have seen nothing in the colony 

 approaching var. ohlongifoUa of Norfolk Island, in which the 

 denticulate leaf is over l^in. long, and I have not had the 

 opportunity of examining Norfolk Island specimens of H. 

 latifolia. The New Zealand plant, which is only known in a 

 fruiting condition, was kindly examined for me by the director 

 of the Royal Gardens, Kew, in 1876, and referred to H. lati- 

 folia var. ; but the differences do not appear to be sufficiently 

 marked to render it worthy of special distinction. Mr. 

 Cheeseman sends a sterile leafy specimen from the Three 

 Kings Islands, characterized by more strict slender branchlets 

 and oblong or oblong-ovate leaves with obscurely sinuate- 

 dentate revolute margins and slender petioles. The leaves 

 are of thinner texture than in H. latifolia, and the reticula- 

 tions on both surfaces are not so strongly marked. 



1. H. crassifolia, Hook. f. Fl. N.Z., i,, 17, t. 8. 



A low rigid spreading shrub, with short stout tortuous 

 branches, bark white, furrowed, branchlets pubescent. Leaves 



