90 INHKKITANCK OF RED COLOUR IN .lUTK. 



red progeny ; similarl}', the impure reds would produce red and 

 green progeny in equal numbers. It may therefore be assumed, 

 that the 1-5 per cent, of red plants found in the green plots, 

 represents about | of the total amount of impurity which would 

 actually have been produced, if all the red plants had been pure. 

 In the latter case, the amount of impurity found might be rather 

 more than 2 per cent. Arguing on this basis, it follows that, 

 under the most favourable circumstances for cross-fertilization, 

 jute would tend to breed true to the extent of about 98 per cent. 

 As a matter of fact, our experience in growing pure single 

 plant cultures of different colours, in adjacent blocks, leads 

 us to believe that the amount of crossing which takes place, 

 under ordinarj^ conditions, between neighbouring fields for 

 instance, is probably much less than 2 per cent. In l'J09, fifty 

 single plants of several different colour types were selected. 

 In 1910, the seed from each plant, sown separately' in adjacent 

 plots, reproduced the parent type absolutely true in ever}' 

 case. Thirty-five out of forty-seven of these plots contained only 

 pure red plants, the remainder being pure green. Immediately 

 adjacent were plots of the F., generation of hybrids already 

 described in this paper. Thus, in the presence of abundance of 

 insects at the time of flowering,* there was no lack of facilities 

 for chance crossing between plants in the green plots and those 

 in the surrounding red ones ; and the percentage of red plants, 

 in the progeny raised from the seed of the green plots, should 

 therefore be an index of the extent to which chance cro.ssing 

 takes place under these circumstances. 



In the present season 1911, seed from these green plots 

 has been sown in xoTi^h acre plots and the number of red plants 

 found in the latter has been counted. The results of the counts, 

 together with the corresponding percentages of red impurity, 

 are given in Table IX. The average number of plants in 

 a 100*'^ ^^^^ P'"*' '^^y ^^ taken as about 1,500. 



• Both it Pii«a and Dacca, insect<i, partioiilarljr ApU florea, TJnv., are abundant vinitors 

 to the 0ower» of Corehorus'eapiutarl*. 



