518 Journal of the Asiatic Society of BengaL [Jnly, 1907, 



Spreng. ; they grow up together, the maize fastest, the pigeon-pea 

 after it and thereby the cotton is held back from maturing 

 until the maize and the pigeon-pea plants have been cut from 

 ofE the land : it is in May when the weather is dry and hot that 

 the cotton harvest b^ns; and it is completed about the middle 

 of June, before the rains break. 



The races of the cotton that make this crop are several : in 

 the course of my work I came across five : — 



(i) Bhogila ; 



(ii) Bara-isar; 

 (iii) Jageria ; 



(iv) Asl Deshi or Bhunchili or Bhuchiii or Bacharia or 



Chtitki; 



(v) Gajar-ganga, 



No. iv is the most widespread : but no. i is that which 

 gives the best outturn. Jageria and Gajar-ganga I only saw 



Mati 



Sarai 



To enumerate here the differences between the races other 

 than in the flower would be to digress. It will be enough to say 

 that accepting Gammie's classification, 



Bhogila j 



Bara-isar > are OossypiuTn neglectum^ var, vera, 



Jageria ) subvar. hengalensis. 



Gajar-ganga ditto, subvar. Kokatia. 



Asl deshi is Gossypiunn intermedium^ 



The drawing above is of a flower of Bhogila, enlarged to 



twice its natural size. The flower is drawn erect, but in nature 



its position is very variable, and perhaps the commonest condition 



is that the flower should face horizontally. The arrow at the 



base indicates one of the five narrow passages by which the honey 



lying between the calyx and the corolla can be drained. That 



lioney accumulates all round inside the calyx, but chiefly below 



the five honey passages. The passages are guarded by short 



hairs, and a proboscis 5 mm. long is required for the reaching of 



the honey. At about 7 a.m. the bright yellow petals become 



unwound and the corolla expands into a funnel ; the anthers at 



the same time dehisce, the mature stigma projecting from 



among them. When the flower opens, the powdery pollen begins 



to fall from the anthers; and it falls generally in chief part 



into the funnel of the corolla ; but any even slight shaking of 



the flower — what with pollen loose in the bell and pollen ready to 



fall from the anthers — causes some to adhere to the stigma. It 



is perfectly evident that the flowers in Behar are every one of 



them early self-pollinated, by the action of the wind, or by the 



settling of insects on the flower outside or inside. A very small 



number of them may be cross-fertilised by insects: the rest 



