142 WILD AND CULTIVATED COTTONS 



plant, so far as I have been able to discover, exists, however, in 

 any herbarium. The plant has long since disappeared from the 

 Calcutta Gardens and, as indicated, no other botanist has recorded 

 the circumstance of its being a wild plant in Ceylon. In the sense 

 advocated by me above it is, however, very largely the plant called 

 G. herbaceum (or G. herbaceum var. obtusifolia) by most writers 

 who have described the cottons of India and Africa. It has, on 

 more than one occasion, been confused with certain forms of 

 G. Nanking, more especially the variety (or race) called rubicunda. 

 Distinc- Briefly, these two species may be separated obtusifolium, the 



tlYG 



points. leaves cordate, the lobes short, more or less ovate-rotund, obtuse or 

 acute, and the flowers yellow; Nanking, the leaves scarcely, if at 

 all, cordate, the lobes oblong, often acuminate, and the flowers 

 purple or with a purple tinge. In Eoxburgh's coloured picture 

 only a side view of the flower is given, but a pink tinge at the 

 base of the tube is doubtless intended to denote the purple spots 

 showing through the yellow. 



Eox- Roxburgh was apparently not aware that it was a wild plant in 



Flora. S Khandesh and Gujarat, nor that it was the type of the most important 

 cottons of Western India. In the Kew Library is preserved a MS. 

 copy of Eoxburgh's ' Flora Indica,' which contains many corrections 

 in his own handwriting, so that it can be accepted as authentic. It 

 furnishes additional particulars, more especially details not given 

 in the work actually published. Under the present species, for 

 example (vol. n. 2067), after a full description, follows the remark : 

 ' It is an ornamental shrub, but the wool is of a very indifferent quality 

 and too scanty to render it an object of cultivation in its present 

 state. It has, however, a good deal of the habit of Gossypium 

 herbaceum, and may probably be the original stock from whence that 

 species and its varieties have sprung.' 



Under the name G. herbacezim, Eoxburgh, in his published work, 

 described the plant now accepted as G. neglectum and linked it with 

 G. Nanking (China and Berar Cottons), but kept all three distinct 

 from G. obtusifolium. 



Deccan In the Kew Herbarium (ex herb. Hooker) there is an Indian 



"tton specimen that I accept as being the plant of Eoxburgh's description 

 and picture. This was collected in the hedges of the Deccan, and 

 bears on the attached label the observation, ' Gossypium scandens 

 vel sylvestre.' (See Plate No. 19, f. B.) So also in Linnaeus' 

 Herbarium (Linnean Society, London) there are three sheets of 



