40 Mr. J. Miers on the Winteracese. 



Dr. Hooker*, after examining the many species of Drimys col- 

 lected in vai'ious parts of the whole American continent, gave 

 his opinion that all of them, including the one from Mexico and 

 that from the island of Juan Fernandez, are mere varieties of 

 D. Winteri. To this opinion I cannot subscribe, for reasons I 

 will here offer ; nor can I agree with St. Hilaire, that the different 

 Brazilian forms which he describes t are only varieties of Z). Gra- 

 natensis. There is certainly a great uniformity of general ap- 

 pearance among the individuals throughout the genus, for, in 

 all, the leaves are quite smooth, thick, entire, oblong, polished, 

 veinless, with an inferior glaucous or whitish surface ; but there 

 is a certain character among them, by which, without mistake, 

 we mav assign to each the country of its origin. There are, 

 however, in each of these several groups so many modifications 

 of form, as well in the leaves as in the inflorescence, that it is 

 almost impossible to offer good specific characters for them. I 

 think it better, therefore, to take the middle course, and regard 

 such individuals as so many varieties, and to consider each group 

 as a distinct species, for which a clear diagnosis may be esta- 

 blished. Under this point of view, the species wdl consist of the 

 New Zealand form, D. axillaris, the Magellanic tree, D. Winteri, 

 the Chilian D. Chilensis, and that from the region of the Ecuador, 

 D. Granatensis, in which I would include the Mexican variety ; 

 I would establish D. Brasilicnsis for the different Brazilian forms 

 enumerated by St. Hilaire, with the exception of D. montana; 

 to these may be added two other distinct species, here proposed, 

 from Southern Brazil ; completing the list by D. Fernandeziana, 

 from the island of Juan Fernandez. 



The younger Linnaeus, who received specimens of D. Grana- 

 tensv< from Mutis, concluded it to be a variety of the Magellanic 

 species J; but Lamarck first pointed out § the forcible distinc- 

 tion between them, and Bonpland subsequently confirmed this 

 by faither observations 1|. These botanists showed that in D. 

 lViat<ri the flowers are always solitary upon several simple pe- 

 duncles, which are aggregated at the termination of the branch- 

 lets ; while D. Granatensis is strongly characterized by each of 

 these peduncles invariably bearing upon its summit three to five 

 long umbellated pedicels, severally supporting a flower. D. Chi- 

 lensis partakes of this latter character, and is therefore equally 

 distinct fi'om the Magellanic species. The general resemblance 

 in the shape and size of the leaves in D. Winteri and D. Gra-^ 

 nutensis is certainly great ; they have also the same dai'k green 

 hue, with a more dealbated under surface, and a similar deep red 



* Flor. Antarct. ii. 229. t PI. Usucllcs, tab. 26, 27, 28. 



X Linii. HI. Suppl. 269. § Diet. Encyc. M^h. ii. 330. 



II Fl. yKfiuin. i. 2()<S, 



