64 BULLETIN 1134, U. S. DEPARTMENT OF AGRICULTURE. 



pansion of the corolla begins. As a result Pima pollen is frequently 

 deposited in upland flowers before the anthers of the latter have 

 opened. 



Insect pollination of cotton at Sacaton. Ariz., is effected principally 

 by Hymenoptera, .the honeybee and wild bees of the genus Melis- 

 sodes being the most important species. Certain Coleoptera and 

 Hemiptera also have been observed to carry pollen, but these in- 

 sects probably are of very minor importance as pollinators. 



By applying the different kinds of pollen separately to different 

 flowers it was sought to ascertain whether different degrees of fertili- 

 zation in Pima cotton result from self-pollination as compared with 

 cross-pollination within the variety and from cross-pollination within 

 the variety as compared with cross-pollination with more foreign 

 pollen. The results were negative so far as concerns cross-pollination 

 within the Egyptian type, there having been no consistent differences 

 in the degree of fertilization attained, whether the stigmas received 

 self pollen, pollen from other plants of the Pima variety, or pollen of 

 another variety of the Egyptian type. On the other hand, somewhat 

 better fertilization of the Pima flowers was obtained in several cases 

 with pollen of a very different type of cotton (upland) than with 

 pollen of the same variety. 



An experiment in which Pima flowers were pollinated, some with 

 Pima and others with upland pollen, gave no satisfactory evidence 

 of a difference in the relative growth rate of the pollen tubes of the 

 two types. 



Selective fertilization, in favor of the related pollen, has been 

 found to occur when pollen of the same variety is in competition 

 with pollen of another Egyptian variety or with pollen of upland 

 cotton on the stigmas of Egyptian cottons. On the other hand, the 

 results of an experiment performed by Argyle McLachlan indicated 

 the absence of a corresponding prepotency of upland pollen on the 

 upland stigmas, and as most of the ovules normally are self-fertilized 

 in both types of cotton, it remains doubtful in what degree selec- 

 tive fertilization contributes to the greater frequency of self- 

 fertilization. 



As both Pima and upland pollen when applied separately to the 

 stigmas of different Pima flowers appear to develop their tubes with 

 equal rapidity and as the upland pollen is at least equally efficient in 

 accomplishing fertilization, it seems necessary, in order to explain 

 the fact of selective fertilization when the two pollens are in direct 

 competition, to assume an inhibiting influence of pollen of the same 

 variety upon the more foreign pollen. 



In localities where bees are not numerous in the cotton fields com- 

 paratively little pollen reaches the upper part of the stigmas of Pima 

 flowers. Under such conditions artificial pollination resulted in a 

 marked increase in fertilization; whereas at Sacaton, Ariz., where 

 the cotton flowers are much frequented by bees, artificial pollination 

 did not affect the degree of fertilization. 



The degree of fertilization resulting from natural pollination has 

 been found to differ in different seasons and at different times in the 

 same season, as well as at different localities. 



Flowers inclosed in paper bags in order to exclude foreign pollen 

 and left to automatic self-pollination are nearly always less well 

 fertilized than open-pollinated flowers, but by artificial pollination 



