42 Oinning Percent of Indian Cottons 



the latter being the sacred perennial cotton only found now in the 

 vicinity of Hindu temples. Lastly G. intentiediuiii Ganimie, a tiinii 

 cultivated as a mixed crop round Allahabad and the West of Bengal, has 

 a ginning percent as low as 15. 



A brief consideration is sufficient to indicate that the ginning percent 

 is not a simple character. It is directly dependent on the weight of 

 seed and weight of lint. In {)ractice the fact becomes still more evident ; 

 thus, among the otfsjjring of nniiierous crosses which have been made 

 between two parent types, each having a gimiing percent of 25 — 26, 

 forms have been isolated for which the figures diverge as much as 

 36 and 18. 



The direction of the present investigation will be most profitably 

 indicated by a short discussion of the a priori considerations on which it 

 was based. From the definition it is clear that the exact figm-e for the 

 ginning percent is dependent on the weight of two distinct bodies — the 

 lint and the seed. This weight is, in each case, dependent on the values 

 of several characters which, in their turn, may vary. Thus, in the case 

 of the seed, the weight depends im the volume, on the specific gravity, 

 and on the number of seeds. In the case of the fibre the characters 

 affecting tiie weight are not so obvious. They may, howe\er, be con- 

 sidered as the weight of the individual fibres — clearly not a unit 

 character as this in its turn depends on such conditions as lengtii, 

 thickness (mean diameter), size of lumen and specific gravity of fibu- 

 wall — and the number of fibres. The relation between the number of 

 seeds and number of fibres in a given .sample i.s again determined by the 

 number of fibres arising fi'om a single seed. The total number of fibres 

 may also be derived fn>m the runnber of fibres arising from a unit ana 

 of seed coat surface — a number tt> wliieh the term "density" will be 

 applied. Calculated in terms of "density " tiie number of fibres will be 

 seen to be goveiiied by the area of seed coat surface which, in its turn, 

 is I'xprt'ssible in terms of volume. Tiiis consideration is of advantage 

 inasmuch as it indicates tiiat the weigiits of both the seed and fibre are 

 directly affected by the same character — the volume of the seed. Whde, 

 however, the weight of seed will vaiy as the cube of the seed radius, the 

 weight of lint will, if the " density " remains constant, vary as the surface 

 area, which varies as the s(|uare of the seed radius. Hence, other factors 

 being constant, the ginning percent will increase as the volume of the 

 seed diminishes. The calculations in terms of " density " are, however, 

 open to objection inasmuch as the number of fibres is probably deter- 

 mined eail}- in the course of development of the ovule while the volume 



