= 
siderable detail, because Professor Van Tieghem and Dr. H. Solereder 
have come to apie different conclusions, working partly with the same 
material. be gathered from the numerous collections cited 
above, Kew eel a very copious set of specimens, every one of 
which I exa e result is the combination of all the 
Chinese specimens under Z. davidiana, from which £, polyandra, 
b. & Zuce., the original species, a native of Japan, is easily distin- 
guished by its mee toothed leaves and uniovulate carpels. £. 
letosperma, o 10ms., a native of Mishmi, North-east of 
Assam, is certai nly very near the Chinese Z. davidiana, Baill., but I 
think it citesirable to follow Dr. Solereder in combining them. The 
imperfect Indian oe are amon | those of a “more robust 
With the permission of Professor Ed. Bureau ‘iid through the kind 
offices = Mr. J. aera I eae been able to examine the actual os 
No. 77, cited by Van tae as belonging respectively to the species 
named. Briefly stated, the anthers of #. Francheti are sometimes even 
shorter than the filaments, and _ carpels are often biovulate and the 
mature ones dispermous. My observations on these and other points 
have been verified for me by Miss M. Smit ai and Dr, O. Stapf. I also find 
that the filaments of some of the stamens in Delavay’s No. 3749 are 
as long as the anthers, and I believe that their relative lengths are t 
some extent due to age. In all the numerous specimens examined the 
number of ovules or seeds, as the case might be, was variable ; in some 
usually one or two, in others usually two o three. 
e accompanying plate was ecg prepared from Merced S speci- 
mens, numbered 1048, and many more drawings were made than are here 
published. Miss Smith opened one  Oeaty in which ihe found ily one 
ovule ; all the others contained either two or three ovules. I have not 
found a single instance of a solitary ovule, but Dr. Stapf, who examined 
several ovaries while I was writing the foregoing, found, in seven ex- 
amined, one uniovulate, four biovulate, and two trievulate. Most of the 
ripe carpels i in Wilson’s specimens are three-seeded. 
I may add that the leaves present no essential differences, and both 
Dr. Henry and Mr. Wilson, who are familiar with Zuwptelea in a living 
state, came independently to the — that all our Chinese speci- 
mens belong to one species.—W. G Hemstey. 
Fig. A, ct nga bogie ng male flowers, —e in —— of four ; B, a bra 
bearing fe ails flowers and young leaves ; C, a branch bearing ripe fruit and rally 
develo: ee leaves. all natural siz 
Fi ract from a male inflorescence; 2s a male flower; 3, a stam 4, 
section Bs a dehisced empty 8 anther; 4, 1 from a male yes 6,a 
emale flower; 7, section of a —* — shen car rpel; 8, section of a young, triovu- 
late carpel, one o aie cy process of a n; 9, a ripe carpel; 10, a section of a ripe, 
one-seeded cz a: a section 0 tek two-seeded carpel; 12, a seed; 13,a 
carpe 
section of the same, ee the post te embedded in the albumen ; embryo. Adi 
enlarged. 
