230 Linnaan Society. 



in length (eight or nine times their width), but with not more thfin 

 eight or nine sets of seeds ; while others again, in their bushy habit 

 and colour, and in the form of their spikes, show an affinity to S. 

 ramosissima. 



All the above-mentioned varieties have oval or oblong seeds, about 

 half as long again as broad, and thinly covered with hooked hairs, 

 upon an even surface. In the two following the seeds are shorter, 

 nearly globular, but covered in the same manner with hooked hairs, 



S. radicans, the next species, is described as differing exceedingly 

 in its mode of growth from any of the foregoing. In all these the 

 root is evidently annual, and produces a single stem, which is hard, 

 and in S. ramosissima may fairly be called woody. In S. radicans, 

 however, a small plant, with only one or two branches, rises at first 

 from the seed. The stem of this lies down, and, generally bur}'ing 

 itself in the mud, sends out radical fibres and new shoots. The 

 process is continued from year to year, the old stems of one year 

 becoming the rhizomes of the next, and these successively dying 

 away as new rhizomes are formed, thus producing a very rambling 

 and diffuse plant. In the preceding forms, every branch and sub- 

 division is terminated by a spike of flowers. In S. radicans many are 

 barren. The spikes, when they occur, are sometimes interrupted, 

 half an inch to an inch long, and composed of about six joints. The 

 colour is a dull greyish green ; with the ends of the spikes brownish, 

 but never red. Though much less abundant than the first, second, 

 and fourth forms, it is by no means rare in the muddy creeks of 

 Sussex and Hants. 



The last form mentioned, under the name -S. lignosa, bears some 

 resemblance in its diffuse mode of growth to S. radicans, and Mr. 

 Woods found some indications of radical fibres from the lower part 

 of the stem, but was unable to ascertain positively the existence of 

 a creeping rhizome. It differs however from S. radicans in the 

 thickness, and firm solid structure of the lower part of the stem, 

 which as in every European species is destitute of annual rings, and 

 attains its thickness and hardness in the course of one year. From 

 S.fruticosa, L., to which it approaches nearly in many respects, it 

 is distinguished by the multitude of its slender branches, and pro- 

 bably also by the structure of its seed, which Koch and Bertoloni 

 describe as tubercled and not hairy in S.fruticosa. The spikes of 

 our English plant are an inch or a little more in length, and about 

 six times their width : those of the true S. fruticosa are usually 

 both absolutely and relatively longer, 



Mr. Woods next makes some observations on the synonymy of the 

 Salicornias described by Ray, who appears originally to have admitted 

 but two species ; the first including all the forms of aS*. herbacea and 

 also S.procumbens; the second attributed by Smith to S.fruticosa, L., 

 but now generally regarded as S. radicans. To these Dillenius adds 

 three others, of which the first, S. myosuroides procumbens, &c., is con- 

 sidered by Mr. Woods as S- radicans ; the second, S. ramosior pro- 

 cumbens, &c., as probably S. procumbens, Sm. ; and the third, S. erecta 

 foliis brevibus cupressiformis , he refers with some doubt to his S. 

 intermedia. 



