NATURE 



241 



THURSDAY, JANUARY 16, 190S 



THE COTTOS PLANT. 

 The Wild and Cultivated Cotton Plants of the World. 

 .1 Revision of the Genus Gossypium. By Sir G. 

 Walt. Pp. .\iv + 4o6. (London : Longmans, Green 

 and Co., 1907.) Price 30X. net. 



SINCE the appearance, in 1S77-8, of Todaro's 

 classic monograph on the genus Gossypium, no 

 serious attempt has been made to deal systematically 

 with the botany of the plants that provide the world's 

 >upplv of cotton. Todaro's work — at once the first 

 ;md last of any practical utility — owes its value to the 

 fact that he worked for the most part with living 

 plants grown by him from seed received from many 

 parts of the world ; the success this method met with 

 justifies the dictum of De Candolle in his " Pro- 

 diomus " when, speaking of this genus, he says, 

 " Hie species a Botanicis admissas recenseam, monens 

 tamen hoc genus monographiae accuratae et ex vivo 

 ■elaboratae maxinie egere." (The italics are not in the 

 original^) Todaro's work requires only to be con- 

 tinued and extended, not to be corrected. 



Anv attempt at a classification of cotton plants that 

 is based on herbarium specimens is doomed to failure 

 ■owing to the impossibility of eliminating the effects 

 that differences of soil and rainfall have in inducing 

 differences of appearance in the leaves and other parts 

 of specimens of the same variety that are not grown 

 tinder identical cofiditions. Many varieties of the 

 genus, even under the most uniform conditions of 

 growth, show a sufticiently wide range of fluctuations 

 to necessitate great care in their determination even 

 \\ hen living plants are being examined ; when we 

 uork on dried specimens, often very fragmentary, 

 grown in widelv distant parts of the world under 

 conditions of rainfall that are not recorded, imaginary 

 differentiations occur to such an extent as to make it 

 impossible to deduce a classification of the varieties 

 that more than indicates broadly the groups into 

 which they fall ; to attempt anything- more than this 

 with herbarium material is to court failure. 



The latest attempt at a monograph of the genus 

 Gossypium is the volume now under review. This, 

 unfortunately, belongs to the category for which suc- 

 cess is an impossibility, being based entirelv — except 

 possibly in the case of a few Indian forms — on 

 herbarium specimens. 



The volume contains some most interesting photo- 

 graphic reproductions of a few type-specimens in the 

 Kew, British Museum and Linnean herbaria, and 

 also (often in colours) of earlv drawings of plants. 

 These reproductions form by far the most valuable 

 portion of the work. 



It is impossible in the space at our disposal to give 

 oven a general impression of the contents of the 

 volume, much less to point out in detail the very 

 numerous errors into which the herbarium method 

 has led its author. No new information is given us, 

 though several new species and more varieties are 

 created on imaginary difTerentiations, while nearly 

 NO. 1994, VOL. 77] 



every variety that has hitherto been described and 

 named receives a new name. 



Thus, of the forty-two varieties mentioned in the 

 volume, Nos. 10 to 25 are cultivated Asiatic or .African 

 forms, and out of these sixteen varieties, classified as 

 belonging to four species, all except the species them- 

 selves receive new names unnecessarily. When we 

 state that many of the synonyms given fail under 

 investigation, and that the descriptions of plants 

 stated to belong to particular species often differ 

 radically from the descriptions given by the authors 

 of the species, it can well be imagined that the volume 

 is rather a retrogression than an advance on the work 

 of Todaro. 



Throughout the book the reader is allowed no oppor- 

 tunity of judging of the correctness of the identifica- 

 tion igiven, since the original definition of the variety 

 is practically never quoted, and must be looked for m 

 the very scattered literature on the genus. It would 

 also have been fairer if, in the case of species here 

 created, a photographic reproduction of the type 

 had been given, and not merely outline drawings, 

 which, throughout the volume, are not good and 

 would not assist in the identification of the plants in 

 the field. 



As an example of erroneous synonymy we may take 

 the following case : — 



G. obtusifolium. Roxb., var. Wightiana. Watt, is 

 stated to be synonymous with G. Wightianum. Tod., 

 and to be the plant constituting the Surtee-Broach 

 growth of India. Now, all botanical considerations 

 apart, Todaro states that the seed that gave rise to 

 his G. Wightianum was as follows :— " Cotone 

 Hingunghatt di Bombay," " Oomrawatt di Bombay," 

 " Howers Barree (sic) di Bombay," " Khandeisk (sic) 

 from .'Vmerican seed di Bombay," " Cotone Hinghun- 

 ghatt Barree (sic)," and " Old DhoUera, provenienza 

 di Bombay." Now, in none of the places mentioned— 

 Hinghunghat, Oomrawati, Barsee, Khandeish, Dhol- 

 lera — will the Surtee-Broach plant grow except in 

 Dhollera; and, curiously enough, Todaro mentions 

 that another sample of Dhollera seed received directly 

 from Bombav gave rise to plants belonging not to the 

 species Wightianuyn, but to the species herbaceum, 

 and giving " un prodotto di bellisima qualita." 



Todaro having already placed the Surtee-Broach 

 plant in the species herbaceum, it would appear un- 

 necessarv that, as in the volume before us, it should 

 be removed into the species obtusifolium, to which 

 it is certainly, to say the least, not more closely 

 related. This removal is further objectionable owing 

 to the great uncertainty as to what plant Roxburgh 

 intended to indicate by the latter name. On this 

 point, again, our author is at once confused and con- 

 fusing; he states (p. 140) that G. obtusifolium, Roxb., 

 is " fairlv common in the hedges of Gujarat," while 

 (p. 135) G. nankin g, var. roji, is also said to become 

 sub-scandent in hedgerows in the same district. The 

 present writer has collected and sown in Gujarat many 

 samples of seed from these hedgerow cottons, and the 

 resulting plants have invariably been identical with 

 one another, and also with the crop known as " roji," 

 which our author classifies as G. nanking, Meyeh, 



M 



