591 



with Mr. Babington, and contend that his rosea has nothing to do 

 with the rosea of Koch and of Duby. The plant I have gathered for 

 rosea in the south of France and east of Germany is a more self-sup- 

 porting plant than most of the genus ; the lower branches indeed 

 spreading, but by no means horizontal and decumbent, as they are in 

 patula and angustifolia, and not unfrequently in erecta. The stems 

 and branches are of an uniform buff colour, instead of green with re- 

 sinous stripes, as in all the preceding species. This seems to me a 

 character of great importance among the Atriplices ; at least within 

 the range of my experience I have not observed it to vary. The ro- 

 sea of Babington is generally prostrate. It is perhaps a good species, 

 though nearly allied to some of the forms of A. patula. I have some 

 suspicion that it is the A. crassifolia of the 'Prodromus.' Koch relies 

 on the indurated and whitened calyx, and the union of the sepals 

 above the middle, for separating rosea and laciniata from the other 

 species. Babington drops these characters, and in his plant the ca- 

 lyx of the fruit is often as dark as it is in patula. 



I have four forms belonging to the section with buff-coloured stem 

 and calyx, which I suspect may be as many species. The first is that 

 of our sea-shores, with the fructification very much among the leaves, 

 the character which Koch gives as distinguishing rosea. The second 

 has long spikes, the upper part of which is naked, the leaves hardly 

 occurring above the branches. This, from the description, I suppose 

 to be the laciniata of Koch and others. The third is the self-sup- 

 porting plant already mentioned. I take it to be rosea of Koch and 

 of Duby, as it certainly is Reichenbach's notion of the alba of 

 Scopoli, usually attributed to rosea. In all these the calyx is nearly 

 square, and more or less tubercled on the back. In the fourth, 

 which occurs with the preceding in the south of France, the calyx 

 spreads from a campanulate base, and is usually without tubercles. I 

 cannot with any confidence apply names to these from preceding au- 

 thors. The first is certainly the English laciniata, and I think also 

 identical with the one authentic specimen of the Linnean herbarium ; 

 but Linnaeus says the stem is erect and the leaves deltoid, which can 

 hardly be said of our plant. The second appears, as I have already 

 said, to be the laciniata of Koch and Deby, and I believe of all mo- 

 dern continental botanists. Perhaps also it may be that of Linnaeus. 

 In both these the fruit is often stalked. The third is, I apprehend, 

 the rosea of Duby and of Koch, but there is in the south of France 

 an intermediate form with naked spikes, the lower leaves of which are 

 hastate and deeply sinuate, while the upper are lanceolate and linear, 



