GABRIELLE L. C. HOWARD. 47 



of the intermediate forms pure, as the extraction of the various 

 intermediate homoz^^gotic forms appears to be the easiest and 

 most conclusive way of determining the principles underlying the 

 inheritance. The range of variation of the parents was deter- 

 mined each year for each character. In 1911, very few of the 

 parent plants were sufficiently well developed to be measured. 

 The cultures were transplanted into a field which had been 

 green- manured. Owing to the exceptionally heavy and pro- 

 longed monsoon of that year, the soil conditions were not 

 favourable to the proper utilization of the green manure, and the 

 plants only developed well in the high-lying portions of the 

 field. The Fo generation of Type 9 x Type 51 and the crosses of 

 this back on to the parents were normally developed, but most 

 of the parent cultures and the F^ generation of the other four 

 crosses were very uneven. This explains why the measurements 

 given for the Fj generation are of a later date than those of the 

 F2. In all these cases the crosses had to be re-made and the 

 work repeated. Similarly, in 1913, as there were a large number 

 of cultures to be grown, some of the lines of the parents were 

 pushed as near the hedge as appeared safe. Unfortunately, 

 the influence of the hedge had been under-estimated, and 

 although the last line was twenty feet from the hedge, the plants 

 developed very slowly, and were so stunted that they had to be 

 rejected. These accidents make the number of measurements 

 carried out on the parents smaller than is desirable. 



I. TIME OF FLOWERING. 



The date of opening of the first flower on each plant was 

 taken as the criterion of this character. All the plants were 

 numbered and the cultures were examined daily at approxi- 

 mately the same time in the morning. The climate of India 

 is very well adapted for all such investigations, as the days are 

 almost invariably sunny, and irregularities due to the periods of 

 cloud and low temperature consequently negligible. The results 

 for the three crosses. Type 9 x Type 51, Type 16 x Type 35 and 



