Dr. Arnott's Notice of the Species of Salvadora. 157 



neously stated by Walpers in his Annales Botanices Systematicse,) is 

 unquestionably a very distinct species, as far as can be understood from 

 the figure, for I have not seen specimens. As to ^S*. oleoides, it is said to 

 be a tree, and to form extensive woods on the western banks of the Jumna 

 and Hyphasis, and to grow also at Khitul and Pallinlah, as well as in 

 the salt sandy parts of the Punjaub, and to be frequent between Agra 

 and Delhi, as far as the desert of Bihassir. Now, from this locality, and 

 the plant being a tree, it appears to me the same as that noticed by Boyle 

 under the name of S. indica; and I find in my herbarium a few fragments 

 of an unnamed Salvadora, collected also on the banks of the Jumna, and 

 inserted by Dr. Wallich in his list under No. 7530, which has the narrow 

 leaves and erect short corolla of Decaisne's plant ; but Jacquemont's 

 species is described and figured with very short (only 3-6 lines long) 

 dense racemes or spikes of flowers, whereas my plant, which, however, is 

 in fruit, has them from 1 to IJ inches long, and neither crowded nor 

 closely flowered ; but in Jacquemont's, as well as in Wall, list. No. 7530, 

 the flower or fruit is almost sessile. 



Of S. perdca, Decaisne remarks that there are a great many varieties; 

 that sometimes the leaves are oval, sometimes linear, and occasionally 

 both kinds are found on the same specimen : in his plant the flowers are 

 distinctly pedicellate, giving the panicled raceme a loose appearance. 



He states, as I have already said, that he only recognises three spe- 

 cies, and yet the Arabic, or at least the Egyptian species, with which he 

 must be well acquainted, does not quite agree with any, unless we are to 

 suppose that he intends to combine all the forms with a reflexed corolla 

 into one. 



I now come to two works published simultaneously within these few 

 months by my friend Dr, Wight at Madras. I allude to the last part of 

 vol. 2 of his Illustrations of Indian Botany, and the last part of vol. 4 of 

 his Icones. In the former a plant is figured at t. 181, under the name 

 oi S. indica; in the latter another at t. 1621, under the name of S. fer- 

 sica, with a portion of a third species called 8. Stocksii ; these three being 

 all the species of the genus which he acknowledges. The first, which he 

 says is common in India, has the corolla reflexed, leaves elliptic-lanceo- 

 late, and the racemes lax ; it is thus identical with the S. j)erdca of all 

 previous writers on Indian botany, Boxburgh and Boyle inclusive. The 

 name of indica is, no doubt, adopted from Boyle, but is certainly not the 

 plant from the Jumna meant by the latter. The second, S. persica of 

 Wight, has narrow leaves precisely as in Boyle's S. indica, or Decaisne's 

 S. oleoides, and like it has the flowers nearly sessile; the racemes, too, 

 are as compact as in Wall, list. No. 7530, to which I have referred, but 

 the corolla and protruded stamens are precisely as in the preceding spe- 

 cies. His third species has the leaves ovate or oval and mucronate, the 

 racemes compact, flowers short-pedicelled, and the corolla is said to be 

 deciduous, while it certainly remains on till the fruit advances to maturity 

 in all the other species. As Dr. Wight hns not poon iho cnvdlln, il is 



