FERTILIZATION IS PIMA COTTON. 3 



such circumstances show but few crosses, indicating that the majority must 

 have been self-fertilized. Judging from the observations thus far made, it 

 would seem that ordinarily only from 5 to 10 per cent of the seeds are nor- 

 mally cross-fecundated. 



Balls, on the basis of his investigations in Egypt, states (6, p. 27) : 



The vast majority of individuals in any cotton crop yet studied are hetero- 

 zygous in several characters, and the amount of crossing which takes place 

 between cotton plants growing in a field so producing this heterozygous condi- 

 tion ranges from 5 to 25 per cent, by experimental evidence. 



The same investigator, in a later publication (7, p. 222), remarks: 

 " In 1905 we found that some 6 to 10 per cent of the ovules in a 

 field of Egyptian cotton were cross-fertilized instead of being 

 selfed;" and he points out that in general culture the apparent per- 

 centage of vicinists is usually larger than the actual percentage, ow- 

 ing to the stronger hybrid plants being retained when the fields are 

 thinned. Certain progenies are mentioned (£, p. 119) in which the 

 percentages of natural hybrids ranged from 25 to 35. 



Allard (2) planted easily distinguishable varieties of upland 

 cotton (Keenan, Okra Leaf, Red Leaf) in alternate rows in northern 

 Georgia and found that progenies grown from at least 20 per cent 

 of the bolls borne by plants of the Keenan variety contained one or 

 more hybrids. Some of the bolls yielded only hybrids, indicating 

 that the flowers from which these bolls developed had produced 

 only abortive or self-sterile rjollen. 



Shoemaker (43) ? working in north-central Texas, found that when 

 plants of the Triumph variety of upland cotton were scattered 

 through a plat of an "okra-leaf" upland strain, so that each 

 Triumph plant was entirely surrounded by plants of the other type, 

 47 per cent of the Triumph bolls, seed from which was planted the 

 following year, yielded hybrids, although these in no case amounted 

 to as much as 50 per cent of the entire progeny of the boll. The 

 proportion of hybrids in the entire population grown from bolls 

 collected on the Triumph plants was 10.9 per cent. No correlation 

 could be observed between the position of the boll on the plant and 

 the extent of the cross-fertilization observed, from which this in- 

 vestigator concluded that " the insects which did the crossing must 

 have worked regularly through the season." 



McLendon (39, pp. 162-167), in Georgia, grew Willett's Red Leai 

 and Hastings Big-Boll (a green-leafed variety of upland cotton) in 

 alternate rows and planted seed from the Hastings plants. In a 

 resulting population of 4,467, 87 (1.9 per cent) of the individuals 

 proved to be vicinists. 



Ricks and Brown (1, pp. 4, 15, 17), in Mississippi, found that 

 when green-leafed varieties of upland cotton were grown in rows 

 alternating with rows of Willett's Red Leaf, the percentages of 

 natural hybrids produced by the resulting seed ranged from 4.9 to 

 11.1. From table 9 of the publication cited (p. 17) it may be deduced 

 that of the bolls borne on plants of Lone Star and of Trice 36 and 

 44 per cent, respectively, gave progenies which contained one or more 

 hybrids with Red Leaf. 



In regard to the prevalence in India of natural cross-fertilization 

 of cotton, Gammie (19, pp. 2, 3), from observations at Poona, con- 

 cluded that it is a very rare occurrence. Evidence to the contrary 



