FERTILIZATION IN PIMA COTTON. 59 



self-fertilization during five successive generations, as compared with 

 continuous open pollination, had resulted in no reduction of the 

 number of ovules. 



Table 37. — Count of the ovules in the 3-celled ovaries of a family of Pima 

 cotton inbred during five generations compared with those of a continuously 

 open-pollinated stock of this variety. 



Stocks compared. Mean. 



Inbred during five generations 21. 0±0. 33 



Continuously open pollinated 20. 8± .27 



Maximum. 



Minimum. 



RATE OF FLOWERING, PERCENTAGE OF BOLL SHEDDING, SIZE OF BOLLS, AND 

 NUMBER, WEIGHT, AND VIABILITY OF THE SEEDS IN AN INBRED POPULATION. 



A family which had resulted from controlled self-fertilization dur- 

 ing seven successive generations was compared in 1921 with a random 

 sample of the continuously open-pollinated commercial stock of the 

 Pima variety, the two populations having been grown in adjacent 

 rows. A record was kept for 88 days of the number of flowers open- 

 ing daily on every plant of both populations, and from these data 

 were compiled the means given in Table 38, which indicate that there 

 was no significant difference in the potential productiveness of the 

 two populations. 



Table 38. — Daily mean number of flowers per plant in a Pima family strictly 

 inbred during seven generations compared with that in a random sample of 

 the continuously open-pollinated stock of this variety. 



Population. 



Number 

 of plants. 



Mean 

 number of 

 flowers per 

 plant,daily 



Inbred 



78 1 1.01±0.02 



Open pollinated 



84 .97+ .02 









Difference 



,04±. 03 







The percentages of bolls shed, the mean numbers of seeds per ma- 

 tured boll, the mean weights of seeds, and the germination percent- 

 ages of the seeds were determined on 15 plants of each population, 

 well-grown individuals which occupied opposite or nearly opposite 

 positions in the two rows having been selected for comparison. Nat- 

 urally pollinated flowers which had opened on the same days on 

 both sets of plants furnished the material. This procedure elimi- 

 nated sources of error which might have arisen from soil heterogeneity 

 or from differences in the weather during the period of development 

 of the flowers and bolls. The data, as given in Table 39 (upper part) , 

 show no significant differences between the two populations except in 

 the mean weight of seeds, in which case the difference was in favor of 

 the inbred population. 



