STl'niF.S IN INDIAN FinRK TI-ANTS. 



left to themselves in thinning this crop always remove any seed- 

 lings which differ from the bulk of the culture. A little extra care 

 in attending to a plot raised for the purpose of growing seed for 

 distribution to cultivators is all tiiat is necessary to keep the cul- 

 ture pure and to prevent any natural crossing. The same care 

 devoted to plots grown for fibre would ensure an absolutely uni- 

 form and even sample. If, as we suspect, it is found that type 3 

 is superior in vigour of growth, in the length and straightness of 

 stem and also in yield of fibre to the other ty])es. the maintenance 

 of this form in a pure condition on a seed farm will not be a 

 dithi'ult matter. If, however, no care is taken to keep the seed-plots 

 pure it will readily cross with dtluM' tv|)es and its (|ualiti('s will l)e 

 quickly lost. 



The significance of the above results are of great im))ortance 

 from the point of view of the improvement of crops in India. In 

 the case of crops which are usually self-fertilized an improved type 

 can often be obtained by the selection of single plants and the com- 

 parison of their offspring in succeeding geneiations. The quality 

 of any such improved strain can be readily maintained bv keeping 

 the seed-plots pure by rogueing and by starting the culture afresh 

 from a single plant when necessary. In attemjiting to apply this 

 method of improvement to crops which cross difficulties at once 

 arise on account of the contamination of the cultures by vicinism 

 before the preliminary work can be completed. If. however, 

 it is possible, as in the case of H. roiuidliiinis, to detect l)y their 

 seedling or early vegetative characters most of the aberrant plants 

 which arise, the difficulties due to crossing can be greatly reduced in 

 the early part of the work and entirelv obviated when one type only 

 is being grown for seed distribution. All that is necessary is to 

 study carefully the seedling and vegetative characters of the select- 

 ed type and to remove systematically everything that a))])ears in 

 any way different from this. All this must l)e com])leted before the 

 first flowers appear .so as to prevent crossing with undesirable types 



It is possible that a study of the .seedling characteis will be of 

 use on cotton seed farms where it is essential to maintain the quality 



