66 CROSS-FEKTILISATION fN THE INDIAN COTTONS. 



and in this the number of individuals is too small to admit of much 

 weight being attached to a percentage figure derived from their 

 use. In all other cases the number of crosses occurring in the early 

 series is proportionately high, for, while the combined percentage 

 is approximately eleven for the " early " series, that for the " late " 

 series is 3"5. Now the flowers from which the seed giving the ' ' early " 

 series was derived were open at the end of the rains, a season when 

 the atmosphere is still moist and the tem[)erature not excessive. 

 The late series, however, is derived from fruit set duiing the intense- 

 ly hot and dry summer months of April and May. In the former 

 case the abundant insect life of the rains is still present, while in 

 the latter the conditions are such as to reduce this to its minimum. 

 It is reasonable, therefore, to conclude that there exists a close 

 interrelation between the abundance of insect life and the amount 

 of cross-fertilisation and the observation adds weight to the 

 opinion already expressed (p. 47) that the divergent views on the 

 occurrence and extent of cross-fertilisation are to be traced to the 

 difference in climatic condition. 



Hitherto the extent to which cross-fertilisation occurs has been 

 considered from the general aspect only. There remains the further 

 problem of the distance at which it may occur. This problem is 

 far more complex, but is also one of great practical importance since 

 the greater the distance at which cottons are found to influence each 

 other, or, in other words, the greater the distance to which pollen 

 is found to be carried, the more intricate must be the precautionary 

 measures taken for extending the culture of new kinds to prevent 

 the swamping effect of accidental cross-fertiHsation. In this con- 

 nection Webber* states that absolute isolation would require a se- 

 paration of from 5 — 10 miles, but that practical isolation may be pro- 

 cured at a distance of a quarter to half a mile. If this proves to be 

 the case with the Indian Cottons the difficulties accompanying any 

 multiplication of a pure race on to a commercial scale will be consi- 

 derable. 



• U, 8. Dept. of Apr. Venr-book. 1902, p. 370. 



