50 



BULLETIN 1134, U. S. DEPARTMENT OF AGRICULTURE. 



haying been bagged. These experiments were performed at Sacaton, 

 Ariz., and in a field near Phoenix where observation had shown that 

 even late in the afternoon the upper portion of the stigmas remained 

 relatively free from pollen. 



The percentages of bolls matured from each lot of flowers and the 

 mean numbers of seeds per boll are stated in Table 28. The data are 

 given for each day separately and for all four dates as one array. 



Table 28. 



-Results of natural and artificial pollination of untagged flowers of 

 Pima cotton at Phoenix and at Sacaton, Ariz., in 1920. 





Naturally pollinated flowers. 



Artificially pollinated flowers. 









1 



Mean number of 







1 



Mean number of 









.a 



seeds 



per— 







,a 



seeds per— 



Locality and 



1 

 i 



o 



I 

 1 

 1 



ercentage of 

 matured. 







i 



I 

 o 



i 



g 



a 

 1 



ercentage of 

 matured. 





date. 



i 



1 

 3 



CO 



si 



| 



! 



CO 



a 



3 



m 



©-. 

 Kg 



i-i 





En 



« 



Ph 



n 



Ph 



B 



fc 



w 



Sacaton, Ariz.: 























Aug. 6.... 



95 



88 



92.7±1.8 



15.1 ±0.21 



1,400±33 



94 



82 



87.2±2.3 



15.1 ±0.24 



1,316±40 



Aug. 13... 



96 



91 



94.8±1.5 



14. 6± .19 



1,385±28 



91 



88 



96. 7 ±1.3 



14. 6± .24 



1,411 ±30 



Aug. 20... 



91 



88 



96.7±L3 



16. 1± .21 



1,556 ±29 



89 



87 



97.8±1.1 



16. 3± .18 



1,595 ±25 



Aug. 26... 



95 



93 



97.9±1.0 



17. 7± .15 



1,734±23 



87 



83 



95.4±1.5 



17. 7± .15 



1,689±30 



All dates.. 



377 



360 



95.6± .7 



15. 9± .11 



1,520 ±15 



361 



340 



94. 2± .8 



15. 9± .11 



1,496±16 



Phoenix, Ariz.: 























Aug. 6 



91 



80 



87.9±2.3 



11.1± .34 



975±39 



98 



93 



94.9±1.5 



15. 8± .23 



1,500±32 



Aug. 13... 



92 



66 



71.8±3.2 



12. 3± .44 



883 ±50 



89 



85 



95.5±1.5 



17. 3± .23 



1,651±34 



Aug. 20... 



72 



62 86.1±2.8 



15. 3± .29 



1,317±50 



95 



72 



75.8±3.0 



16. 7± .22 



1,266±53 



Aug. 26. . . 



89 



76 85.4±2.5 



17. 6± .23 



1,501 ±48 



94 



90 



95.8±1.4 



17. 9± .17 



1,716±3Q 



All dates. . 



344 



284 82.6±1.4 



1 



14. 0± .20 



1,157±26 



376 



340 



90.3±1.0 



16. 9± .10 



1,526±19 



1 The probable error of this value was computed by the formula V(Ab) 2 +(Ba) 2 , A±a being the per- 

 centage of bolls matured and B±b being the mean number of seeds per boll. 



Considering the combined results for the four dates, it appears 

 that the fertilization of the naturally pollinated flowers at Phoenix 

 was significantly inferior to that at Sacaton, the difference in the 

 percentage of bolls matured having been 13 ±1.6 and the difference 

 in the mean number of seeds per boll having been 1.9 ±0.23. At 

 Sacaton it is evident that artificial pollination did not result in more 

 nearly complete fertilization than was attained by natural pollina- 

 tion, neither the percentage of bolls matured nor the mean number 

 of seeds per boll having differed significantly in the two lots of 

 flowers. Artificial pollination at Phoenix, on the contrary, signifi- 

 cantly increased the degree of fertilization, the increases over the 

 results from naturally pollinated flowers having been for the entire 

 period 7.7±1.7 in the percentage of bolls matured and 2.9±22 in the 

 mean number of seeds per boll. In the mean number of seeds per 

 100 flowers, a value which integrates the percentage of bolls matured 

 and the mean number of seeds per boll, the increase due to artificial 

 pollination amounted to 32 per cent, indicating that a substantially 

 greater crop both of seed and of fiber 29 might be expected if bees 

 were abundant in the Salt River Valley cotton fields during the 

 blossoming period. 



29 Evidence that the weight of fiber per boll 

 been presented elsewhere (28). 



correlated with the number of se^'ds has. 



