MB. O. BENTHAM ON TEBNSTB(EMIA.CEiE. 55 



quite meet in the axis, and even diverge at the summit, yet they 

 are not so decidedly apocarpous as in almost all Dilleniacece. Dil- 

 lenia itself is, it is true, anomalous in the Order, having almost the 

 carpels oi Actinidia, and no arillus to the seed; yet the habit, the 

 anthers, the shape of the seed, the albumen and embryo are those 

 of other Dilleniacece, and very different from those of Actinidia 

 and Saurauja. In several species also of Ternstroemia, Adinandra, 

 and other Ternstroemiaceous genera, the carpels, although con- 

 verging and uniting at the base of the styles, are more or less 

 distant lower down, leaving a hollow centre. InSaurauja the gamo- 

 petalous corolla, the insertion of the stamens, the styles, and many 

 other points, are so decidedly Ternstroemiaceous as to prevail over 

 the Dilleniaceous foliage, the embryonic character being in this in- 

 stance one of degree only. We have proposed, therefore, to retain 

 both Actinidia and Saurauja in Ternstroemiacecs, although in a small 

 separate tribe, distinguished as above from Ternstraemice proper. 



Of the eight species known of Actinidia, two have the sepals so 

 little imbricated that one had been referred by Gardner to 

 Tiliacece (under the name of Heptaca); but the other species 

 show that the aestivation is essentially imbricate, although less so 

 than in the generality of Ternstroemiacecs* : . 



To the two genera Saurauja and Actinidia we propose to add, in 

 the same tribe of Sauraujece, Zuccarini's Japanese genus Stacliyurus, 

 of which we have a second species, gathered by Dr. Hooker in 

 Sikkimf. Originally referred to Pittosporece, it has been very pro- 



* Zuccarini published five species of Troclwstigma, of which we have not seen 

 authentic specimens ; but there is reason to believe that two of them (T. arpdum 

 and T. ritfum) are but slight varieties of the K Indian Actinidia callosa, which 

 certainly includes A. Kolomikta, Eiippel in Maxim. Fl. Amer. 63, from the Amur 

 Eiver, and has been gathered by Wilford in Mantchuria. If such be the case, 

 we have, besides these four Japanese species, two from South China (A. chinen- 

 sis, Planch., and A. Championi, Benth.) and the two following new ones :— 

 y^yr A. striffosa, Hook. fil. ct Thorns. Dioica?, strigoso-hispida, foliis pctiolatis, 

 ovatis oblongisvo acuminatis, calloso-ciliatis denticulatisve, utrinque viridibus, 

 peduncidis brevibus 2-1-floris, calycis glabriusculi lobis imbricatis — Flores 

 majusculi. Styli subulati, divergentcs, vix apice dilatati. 



Sab. Sikkim, altit. 6-8000 feet, J. D. Hooker. 



A. eriantha, Benth. Foliis brevipetiolatis ovali-oblongis breviter acuminatis 

 minute ealloso-dentatis supra glabris subtus ramulisque dense cano-tomentosis, 

 pedunculis brevibus paucifloris, calycis tomentoso-lanati lobis vix imbricatis — 

 Folia 3-4-polliearia, petiolo lanato 3-4-lineari. Flores majusculi. 



Rah. S. China. Communicated by Dr. Lindley. 



t Stachyurus hhnalaicus, Hook. fil. et Thorns. S. Japonic* sumlhmus, diflert 

 foliis brevius pctiolatis longioribus longe acuminatis, baccis subsessdibns sub- 



globosis obtusis. 



i- o 



