SOUTH AMERICAN PLANTS. 89 



leaves, that they are frequently confounded together*. For 

 these reasons 1 have been led to search for more certain specific 

 characters in the structure of the flowers, which appear to afford 

 constant features that may be relied upon, and this has induced 

 me to remodel the genus, and revise all the species within my 

 reach. As Lycium has recently been so fully elaborated by M. 

 Dunal, and the numerous species described by him have been 

 accompanied with such copious and minute details, this may ap- 

 pear to be quite unnecessary; but my inquiries have convinced me, 

 that for the purposes of specific distinction, little value is to be 

 placed upon many of these ample definitions, and that it is re- 

 quisite to examine the same materials again with more caution. 

 M. Dunal enumerates only three South American species ; I have 

 here described thirty : during my travels thirty years ago, I met, 

 on the confines of the Andes, with many plants hitherto un- 

 noticed; and I find in Sir W, Hooker's rich herbarium, other 

 new species, besides several from the northern portion of that 

 hemisphere, as well as many of Asiatic and African origin, which 

 I now propose to describe* 



Of the 70 species I have enumerated (besides those that are 

 dubious), 33 belong to the old, and 37 to the new w^orld. Of 

 these, 3 are found in Eui-ope, 2 in Madeira and Barbaiy, 3 in 

 Tartary, 6 in Arabia, Persia, Guzzcrat and Scinde, and 19 in 

 South Africa; and in the other hemisphere, 1 in the United 

 States, 6 in California and Mexico, 1 in the West Indies, 2 in 

 Peru, 6 in Chile^ 3 in Southern Patagonia, 13 in the extensive 

 shingle plains that skirt the eastern flanks of the Cordillera, or 

 that penetrate its gorges, 3 in the vast mud deposit that forms 



* This was long ago (1813) showu by Pohet, who says. Diet. Method. 

 Suppl. iii. 427 : — " L^ plupart des especes qui composent les Lycium sent 

 tellement rapprochees, qu'elles sont difficiles a bien caracteriser, d'autant 

 plus que la plupart cultivees, varient dans la forme ct la grandeur de leur 

 feuilles, dans le nombre des divisions du calice et de la corolle, d'oii il est 

 resulte de la confusion dans la synonymic, et beaucoup de doutes ])our 

 quelques especes, etablies par des auteurs moderncs." Indeed the habit of 

 Lycium is peculiar, where we observe a constant tendency to the abortion 

 of its branchlets, especiallj^ when grown in arid and saline places; we then 

 commonly find in each axil of the stems, a protuberant knotty excrescence, 

 sometimes quite bare, which is, in fact, a leaf-bud or gemma, checked in 

 its earliest development and lignified; generally, a few elementary leaf- 

 scales succeed in escaping from the gemma, forming a fascicle of staned 

 leaves, and often the node is at the same time expanded into a longer or 

 shorter spine, which again bears several similar suppressed nodes. In the 

 same plant, however, when exposed to circumstances that favour a more 

 rapid growth, we observe the nodes expanded into regular lengthened 

 branches and branchlets, with much larger alternate leaves, in each axil. 

 Hence it is, that the appearance of each species may become, and usually 

 is so varied, that we are liable to constant error in determining its indivi- 

 duality from its habit alone, and the form or size of its leaves or sj)ine8, 



N 



VOL. II. 



