IO2 CEREALS 



random, thus giving in all about 100 cultures to compare 

 in the second season. The object of this was to deter- 

 mine whether by so spacing the plants " yield " quality in 

 a parent plant could be detected by the eye. The results 

 have been negative. Many of the apparently superior 

 plants of the first year have proved in the second year to be 

 splitting cultures due to natural cross-fertilization, their 

 exuberance of growth and tillering in the first season 

 being simply the well-known phenomenon associated with 

 all first crosses. Hence this preliminary work has been 

 abandoned, and in future comparison will be made directly 

 between the lines. At the moment of writing the results 

 are not available, nor has the soil error been determined. 

 The determination of the latter for both nursery and 

 transplanted field is an essential to accurate work of this 

 kind, but is only made possible when a sufficiently large 

 stock of a single pure line culture has been obtained. It 

 is hoped that this determination will be completed during 

 the present season. 



When the degree of probable error has been determined 

 it will be possible to say whether intra-varietal selection 

 for "yield" is an easy possibility or not. So far as is 

 known, the literature of rice cultivation gives no definite 

 data on this question. It abounds in varietal tests and 

 trials in which the variety is taken as the unit and com- 

 pared with other varieties. The question, however, 

 which concerns us at present is whether a variety which 

 is specially adapted to the trade requirements in the 

 character and quality of its grain, but of low yielding- 

 power, can by selection among its lines be transformed 

 into a type which will satisfy the demands of the agricul- 

 turist as well. 



