FERTILIZATION IN PIMA COTTON. 49 



pollen of a very different type (Egyptian) competes on nearly equal 

 terms with pollen of the same variety. 27 



RELATIVE COMPLETENESS OF INSECT POLLINATION AT 

 DIFFERENT LOCALITIES. 



Observation in Arizona has shown that the number of efficient 

 pollinating insects differs greatly in different localities. 28 Bees and 

 other active pollinators are normally abundant among the cotton 

 flowers at Sacaton throughout the summer, and the entire surface of 

 the stigmas is almost invariably well covered with pollen soon after 

 the corolla has opened. On the other hand, observations in the 

 Salt River Valley, at distances of 25 to 40 miles from Sacaton, have 

 shown that insect pollination of cotton there is often much less rapid 

 and complete. The probable explanation is that in recent years an 

 extensive and almost continuous acreage has been planted to cotton, 

 and the insect population is not large enough to insure thorough 

 pollination of all the flowers. 



Thus, on July 18, 1919, in a field situated near Tempe in the heart 

 of the cotton-growing district, no pollen grains were observed upon 

 the extrastaminal portion of the stigmas at 9 a. m. and very few at 

 10 a. m. Late in the afternoon of July 20, 1920, inspection of the 

 same field showed the extrastaminal portion of the stigmas to be 

 free from pollen in most of the flowers, while the remainder bore 

 only a few insect-transported grains. None of the flowers examined 

 showed thorough pollination of the whole stigmatic surface. Two 

 other centrally located fields, one at Phoenix and one near Tempe, 

 which were examined at 5 p. m. on August 5 and at 4 p. m. on August 

 6, showed similarly deficient pollination. On the other hand, in 

 fields situated on the outskirts of the valley, at Litchfield and at 

 Goodyear, which were examined at noon on the same days, bees and 

 other pollinators were abundant, and the stigmas of the cotton 

 flowers were found to be well covered with- pollen. 



Experiments were made in 1920 with the object of comparing 

 the relative degree of fertilization by natural pollination in fields 

 where observation had shown, on the one hand, thorough pollination 

 of the entire stigmatic surface and, on the other hand, a deficiency of 

 pollen on the upper portion of the stigmas. It was sought also to 

 ascertain whether in the latter case artificial pollination would in- 

 crease the degree of fertilization, as compared with that of naturally 

 pollinated flowers. 



At weekly intervals during the month of August the stigmas of 

 approximately 100 flowers were smeared with pollen from other 

 Pima plants, and an approximately equal number of flowers were 

 tagged and left to natural pollination. The only difference between 

 the two treatments was the thorough cross-pollination of the entire 

 stigmatic surface of the artificially pollinated flowers, neither lot 



"Attempts made at Sacaton, Ariz., to determine whether selective fertilization occurs 

 in upland cotton have been unsuccessful, owing to the loss by shedding of nearly all 

 of the bolls from the treated flowers. 



28 A pronounced difference in the abundance of pollinating insects at different localities 

 in Arizona was noted by Cook, McLachlan, and Meade (12, p. 34) : "At the time of our 

 visits to the fields at Yuma and Sacaton there was a notable difference in the activity 

 of the insects at the two places. Several species of large wild bees that were industri- 

 ously visiting the flowers at Yuma in September were not seen at all at Sacaton." 



