40 rROSS-FKRTIMSATION I\ TlIF, IXPIW COTTON'S. 



rilo or \vhi<'li oxhihit a lack of vogotativo vigour. Tho subject, 

 therefore, may be best handled by tracing the progeny of indivi- 

 dual plants through a series of generations. 



Observations on thk incidence of sterility. 



Tijpe 3.* G. arborenm, L. The Niirmn cotton of Northern 

 India. In 1906 three plants of this type were grown and from 

 these seed was obtained by the simple ])rocess of bagging the 

 flowers. Of seventeen flowers so bagged eight set and gave seed. 

 Of the plants raisetl from this seed in 1907, twenty flowers were 

 similarly bagged but of these three only set. Later in the season, 

 therefore, a further series of six flowers, in which cross-fertilisa- 

 tion from sister plants was eflfected. was bagged and of these two 

 set. Mainly owing to the fact that Nurma cotton is readily ob- 

 tained in a state of purity as the result of its growth being confined 

 to the neighbourhood of temples and at a distance from the ordi- 

 nary field cotton, and also to the fact that it is a perennial and can 

 be rattooned, no further attempt has been made to grow this type 

 from seed derived from self-fertilised flowers. The numbers are 

 small but, as far as they go, show a marked diminution in 

 fertiUty as measured by the percentage of fruits which set in the 

 second generation. 



Type 9. A form of (1. neqlectum. Tod., with white flciwer and 



a leaf-factor y3. A .single plant culture of this ty])e has been 



maintained since 190G. The original ])laiit was then chosen on 

 account of its prolific habit, the weight of bolls necessitating the 

 use of artificial support. In this year fourteen flowers were pro- 

 tected and seed was obtained from seven of these. Of the off- 

 spring twenty-one flowers were similarly protected in the follow- 

 ing year and of these nine only set. In 1908 some sixty plants 

 were raised from the seed so obtained and of ten floweis protected 

 not one set. In this case the carpels of the protected floweis did 



• For th« numbers here used to deiiore llic types (/. Leakr-, Jour, of (iL-iietics, I, |it. 3 

 (191 1), an<l apiiendi-i to tin's note, 



