236 KUMPTA COTTON AND ITS IMPROVEMENT 



The curve, it wdil be seen, shows two points of very great frequency, one 

 where the plants have three monopodia, the other where there are five, the 

 latter being a little the more common. The mean number of monopodia 

 is 4-15. 



Such a multimodal curve means in this case, probably, that there are a 

 number of different strains of cotton intermixed, with different tendencies to 

 form monopodia. This probability is made almost into a certamty by a 

 consideration of the variation in other characters. 



(2) The shape of the lobes of the leaf. The lobing of the leaves is often 

 regarded as a useful character in classifying cottons, and in most descriptions 

 of species and varieties the shape and size of the lobes is usually referred to- 

 Leake^ was the first, however, to introduce any system of recording this 

 character by measurement, and his 'leaf factor' was in.tended to supply the 

 necessary accurate expression of the depth of the leaf indentation combined 

 \Adth the Avidth of the lobes. It consisted in the maximum length of the leaf 

 from the tip of the central leaf lobe to the base of the petiole (1), minus the 

 distance from the bottom of the adjoining indentation to the base of the 

 petiole (1^), divided by the maximum breadth of the central lobe (b). It is 

 mdicated, in fact, by the expression -^ • This, as Leake remarks, is a purely 

 empirical figure, but nevertheless, affords a useful means of indicating the 

 shape of the leaf. I have obtained almost the same figm-e by sunply 

 measuring the length of the central lobe of the leaf and dividing it by the 

 maximum breadth of the lobe. 



Certam precautions are obviously necessary in the case of all sueh 

 measurements. A primary leaf on the main stem differs much from the 

 secondary leaf on the branches of the same plant. The primary leaves on any 

 one plant, however, vary very little with the exception of the first three. These 

 are in many cases abnormal, sho\\Tng a smaller number of divisions than the 

 remaining leaves on the plant. To secure uniformity of method, however, all 

 measurements were taken on the sixth, seventh, or eighth leaf from the base 

 of the stem. 



Before oiving the actual figm-es obtained in measuring the length and 

 breadth of the middle lobes of such leaves of kmnpta cotton, I may say that a 

 simple experiment in crossuig between two pm'e lines, breeding true, which 

 I had isolated, indicates that mthe length and breadth of the lobes of the 

 leaf we are dealuig with two simple Mendelian factors. It was found, in 



1 Jownal of the Asiatic Society of Bengal, Vol. IV, No. 1 : Journal of Genetics, Vol. I, No. 3. 



