330 ON RUBUS RAMOSUS.
to know whether this discrepancy has been noticed by other observers in this or in other species. In E. Centanrium and its subspecies the stamens have a very peculiar habit of not surrounding the pistil, bnt, when they are discharging their pollen, the anthers being all collected together on one side of the stigmas. In Sjnr<2a FlUpendida the stamens themselves are so arranged as to favour cross-fertilization. The outer rows of sta- mens are matured first, and while they are discharging their pollen, the inner immature stamens are folded up into a kind of cap, completely covering up the stigma and preventing the access of pollen.
The following comprises all the additions I am able to make this year to the three lists: —
Protandrous. Synacmic. Protogynous.
Silene maritima. Hj^ericum pulchrum. Lonicera Periclymenum.
Spiiaua Filipendula. Galium palustre. Plaatago maritima.
Hedera Helix. Hieracium umbellatum.
Centaurea nigra. Convolvidus sepium.
Sen'atula tinctoria. Veronica Chamaidrys.
Solidago Virgaurea. Prunella vulgaris.
Jasione montana Galeopsis Tetrah.it.
Melampyrum pratense.
Lysimachia nemorum.
In many Composifce belonging to the suborder TubnUflorere, the anthers belonging to the perfect bisexual flowers of the disk, while protandrous as respects their own pistil, are developed at the same time as the stigmas of the female flowers of the ray.
I do not know wdiether it may have been noticed that in several species of Couvolviihice't the stamens are normally of unequal lengths, as if indicating a structural approximation to the didynamous Orders Lahiatee or Scropludarinere . Jn Convolvulus arvensis the filaments appear to be never quite ecpial in length. I almost invariably find two longer than the other three; the anthers of one of the longest frequently resting on the fork of the mature bifid stigma at the time that it is dischargnig its pollen. In C. sepium the filaments are usually of precisely the same length; but in C. tricolor, L. (C. minor, Hort.), and Pharbitis Jdspida, Chois. (^Convolvulus major, Hort.), the inecpiality -is quite as strongly marked. As is likely to occur in cultivated plants, the variation is in these cases subject to greater irregularity. In the former species I have sometimes noticed the diffe- rence in length to be quite as great as in most didynamous plants; in the latter species there appear often to be three distinct lengths, viz. two long, one medium, and two short.
��ON RUBUS RAMOSUS, Blox., AN UNDESCRIBED SPECIES OF THE NUDICAULIS GROUP.
By T. R. Archer Briggs.
Through the kindness of Mr. J. G. Baker, I have recently had the opportunity of comparing many of our Plymouth Rubi with named examples of foreign specimens in his herbarium from Genevier, Mercier, Wirtgen, and other Continental students of the genus. One comparison that I have made has been between the Schlickund of Wirtijen and a
�� �
