WOODHOUSE AND TAYLOR. Ill 



many as six insects being found together on one plant in the 

 mornings. Lepidoptera were represented by a few specimens of 

 Parnara mathias, F. ; Pachyzancla stultalis, Wik, and Zinckenia 

 fascialis, Cram. A chrysomelid beetle and Epilachna dode- 

 costigma, Muls., were common on the leaves of the plants. 

 Several muscid flies were almost as plentiful as the bees to 

 which their behaviour was similar. It is probable that bees and 

 Hies sometimes cross-fertilize some of the younger flowers. 



3. Extent of Natural Crossing. — Messrs. Piper and Morse's 

 observations at the Arlington Experiment Station have made 

 them conclude that even when the test rows are grown contio-u- 

 ously, the percentage of hybrids is perhaps not one individual 

 in two hundred. They also find that the hybrid seed can be 

 recognised by the presence of peculiar markings on the seed. 

 Heterozygote plants can be distinguished by the appearance of 

 the pods at the top of the branches, which pods are more tumid 

 less hairy and of thinner texture. Our observations at Sabour 

 lead us to conclude that natural crosses do not occur on the 

 plains of India to such an extent as in America. The three 

 varieties which have been grown next to next since 1909 were 

 found to be pure for all characters in that year and have since 

 retained those characters. In 1910 eleven single plant cultures 

 were grown of the three types I, III & IV, but no rogues were 



d among them. In 1911 sixty-four plots derived from 

 single plants in 1909 or 1910 were grown in lines next to next 

 and of these only one plot was found to be a hybrid. This plot 

 (E 420 of II) was grown from the seed of a single plant of 

 the yellow variety (type III), which was selected on account 

 of its greener colour and darker hilum. Isolated rogues were 

 also found in a few other plots and will be tested in 1912 

 It will probably be found that most of them are chance 

 admixtures. A statement of the rogues found in 1911 is given 

 below. In all, 3,587 plants were harvested from unbagged seed 

 grown from single plants in 1910. Column 3 gives the number 

 of years during which each single plant culture has been 



