Linnaan Society. 231 



Then follow some remarks on the characters of Arthrocnemum, a 

 genus separated by M. Moquin-Tandon from Salicornia, principally 

 on account of the different form of its embryo, and to which he refers 

 S '. fruticosa and S. radicans. In all specimens of 5. radicans, and 

 in some of what is called S. fruticosa, Mr. Woods finds the seeds 

 apparently destitute of albumen, and with the radicle lying against 

 the edges of the cotyledons; but in the true S. fruticosa, supposing 

 that name to be correctly applied only where the seed is tubercled 

 and hairless, he finds a portion of albumen, but the extremity of the 

 cotyledons still close to the point of the embryo. 



The author concludes with the following resume : — " If I were 

 to sum up the result of my observations of this year on the genus 

 Salicornia, I should say that S. procumbens is a distinct species ; that 

 S. radicans and S. lignosa are certainly specifically distinct from S. 

 herbacea ; but whether they are so from each other, and whether, if 

 that be the case, S. lignosa ought not to be considered as a variety 

 of S. fruticosa, L., and the plant with tubercled seeds to be called 

 S. megastachya, I do not feel competent to decide. The other forms 

 of S. pusilla, S. intermedia and S. ramosissima, may perhaps be 

 varieties of S. herbacea, but this also is a subject for further investi- 

 gation." 



The paper was accompanied by specimens of the various forms 

 therein described, for the Society's herbarium ; and a note was added 

 by Air. Kippist, Libr. L.S., who at the request of Mr. Woods had 

 examined the seeds of the specimens sent. He had found the struc- 

 ture of the embryo to be nearly the same in all the British forms ; 

 consisting of thick, fleshy, almost semicylindrical, bright green coty- 

 ledons, in some species scarcely wider than the radicle, which is bent 

 sharply round, and lies not against their edges, but on the back of 

 one of them, the radicle being therefore incumbent. In one or two 

 instances the cotyledons were found to be inclined rather obliquely 

 towards the radicle, but this appeared to be the result of accidental 

 pressure, the majority of the seeds examined of each variety pre- 

 senting the same character of incumbent cotyledons. In all, the 

 albumen was either entirely wanting, or in very small quantity. This 

 seems to be equally the case with S. radicans, notwithstanding that 

 M. Moquin-Tandon refers this species, as a variety, to his Arthro- 

 cnemum fruticosum, to which he ascribes copious albumen. As re- 

 gards the structure of the seed, Mr. Kippist agrees with Mr. Woods 

 in thinking that the S. radicans, Sm., would be much better placed 

 in Salicornia, as defined by Moquin-Tandon, than in his genus Ar- 

 throcnemum, to which he attributes a crustaceous testa and semi- 

 annular peripherical embryo, characters which Mr. Kippist had not 

 met with in any British species. In all the specimens gathered by 

 Mr. Woods on our own coast, the covering of the seed is thin and 

 membranous, and clothed with hairs, which differ much in length in 

 different species. In S. herbacea and the species most nearly related 

 to it, they are of a sigmoid form, spreading at the base, but curled 

 inwards at their extremity, unbranched, and destitute of septa or 

 spiral fibre. They are longest in the form which Mr. Woods calls 

 intermedia, much shorter in lignosa, while in radicans, they are so 



