РТ. 
296 MR. S. T. DUNN : А REVISION 
imbricate, in fact almost valvate, the sepals of the male flowers of that species 
are. Gardner had no female flowers. Probably Champion, who sent him the 
specimens from Hongkong, was unable to find that sex, as it is extremely rare 
there, though the male plants are frequent about the island. The true affinities 
of the plant were recognised by Bentham in his * Flora Hongkongensis? a few 
vears later. 
Meanwhile, a sixth species, known in Manchuria as Aalomikta, was described 
by Maximowiez * as а Prunus and was subsequently made by Regel f the 
type of a new genus under the name of Aolomikta, being finally placed in its 
true position as Actinidia Kolomicta by Maximowiez. A few years later 
Hooker and Thompson added a species from India f and Bentham another 
from S. China $. The latter was the first example of the type, now well 
known, which is distinguished by the thick indumentum of the leaves and 
flowers. Meanwhile important collections From the interior of China were 
beginning to arrive in Europe, and between 1884 and 1895 five new Chinese 
members were added to the now considerable genus by Hemsley ||, Franchet|, 
Maximowiez **, and Hance ff. To these thirteen species, seven more were 
added between 1905 and 1909, all from the interior of China, while the 
novelties described in this paper bring the total number up to twenty-four. 
Systematic Position.—lt may be stated by way of preface that practically 
all botanists who have given attention to the systematice position of Actinidia 
have placed it either in Dilleniaceæ or in Ternstræmiaces, and that it is there- 
fore unnecessary to discuss its relations with any other families. The founder 
of the genus considered 22 that it belonged to Dilleniaceæ in consequence of 
its numerous carpels, free styles, and, as he supposed, innate anthers, all of 
which were characteristic of his idea of that Order. Their absence among 
the genera, so far as they were known to him, of Ternstræmiaceæ would have 
prevented him from including it in that family. Не did not possess fruit. 
When six years later Siebold and Zucearini $$ with abundant material 
including fruit described their genus 7rochostigma, which, as we have seen, 
proved to be congenerie with Actinidia, they recognised its close affinities 
with баитаца and placed it, but with some doubt, in Ternstremiacem. In the 
same year, 1843, it was included in the third Supplement of Endlicher’s 
‘Genera Plantarum | |; and placed next to Saurauja in his tribe Sauraujez of 
Ternstromiaceæ. The next two Actinidias that were discovered, happening to 
have their sepals only slightly imbricate, and their fruit not being available, 
* [In Rupr. in Bull. Phys. Май. Acad. ху. (1857) 262, 
+ Bull. Phys. Math. Acad. ху. (1957) 219, А. strigosa. 
$ A. eriantha. | A. lunata. 4 1. Davidii and A. melanandra. 
o 4. tetramera. TT A. fulvicoma. 
it Lindley, Introd. Nat. Syst. ed. и. (1836) 459. 
$$ Abh. Akad. Wiss, Muench. iii (1845) 726. | Page 94. 
