Leake and ram prasad. 47 



but it is a significant fact that tlie only types grown pure are those 

 whose natural habit helps to preclude the occurrence of cross- 

 fertilisation. 



(c) Previous observations on the subject. 



From time to time the most divergent views on the extent to 

 which cross-fertilisation takes place in the field have been advanced. 

 At one extreme lie the views of Gammie who records* a series of 

 observations which led him to conclude that ' ' Indian Cottons are 

 normally self-fertilised." At the other Cookf considers that it 

 occurs with a frequency that might render it possible to produce on 

 a commercial scale Fj crosses by the simple process of growing the 

 two kinds it is desired to cross in alternate lines. The crop raised 

 from the seed so obtained will contain so many crosses that the pure 

 parents may be rogued out without excessive thinning of the crop. 

 It will be noted that these extreme opinions are based on observa- 

 tion of very different types growing in very distant localities. 

 There is, therefore, no inherent difficulty in believing that both opin- 

 ions may be correct for the particidar types and for the particular 

 area in which the observations were made. The structural similar- 

 ity of the flower in all the types is, however, very great and on a 

 priori grounds, therefore, it would seem probable that such differences 

 as occur would be shght. It has been shown above that such cross- 

 ing as takes place is caused by insect visitors and any divergence 

 found in the amount which occurs is to be sought, therefore, not 

 in the structural peculiarity of the flower but in the climatic condi- 

 tions which obtain when the cotton plant is in flower and which 

 directly affect the abundance or paucity of insect life. 



Of the expressed views which lie between these two extremes 

 those which are the result of observations made in India on the In- 

 dian cottons are most germane to the present question and may be 

 dealt with fir-st. It is impossible here to detail all the earlier obser- 

 vations on the subject. Reference to these is to be found in Watt's 



* The lodian Cottons, I'JUO. 



f U. S. Dept. of Agr, Bureau Plant Industry Cull, 147. 



