38 REPORT OF THE AGRICULTURAL RESEARCH INSTITUTE 



vation. A considerable volume of results has been obtained 

 which has now been arranged for publication. As in Java 

 indigo, the well-being of the crop depends to a very large 

 extent on the physical condition of the soil and on a 

 copious supply of air to the nodules and roots. The results 

 obtained, on different classes of soil in the Botanical area, 

 explain both the present geographical distribution of gram 

 in India as well as the dependence of the yield of seed on 

 the season. The two chief climatic conditions which limit 

 the yield are heavy rains, which produce surface crusts and 

 deprive the roots of air, and damp weather at flowering 

 time which interferes with pollination. Self-pollination 

 is the rule in gram at Pusa but instances of natural cross- 

 ing occasionally occur. Twenty-five types, differing widely 

 in habit and yielding-power, have been isolated which are 

 now being tested for yield under varying conditions. One 

 interesting fact has already come out of these variety trials, 

 namely, the union of high yield and good quality in the 

 same variety. Type 9, grown at Pusa for the last four years 

 on a large scale, on widely varying soils and in very differ- 

 ent seasons, has given an average yield of just over twenty 

 maunds per acre. This is the second highest average out- 

 turn, the best being that of twenty maunds thirty-three 

 seers yielded by Type 18. Taking both yield and quality 

 into consideration, however, and on the basis of the valua- 

 tion of Messrs. Ralli Brothers, Type 9 gave the best return, 

 an average of Us. 78-11-0 per acre while the average pro- 

 duce of Type 18, the highest yielder, was worth only 

 Rs. 61-4-0 per acre. This result is another illustration of 

 the value of selection methods in improving Indian crops in 

 the present condition of agriculture in this country. Had 

 an attempt been made by hybridization methods to achieve 

 such a union of yielding power and grain quality, the work 

 entailed would have been arduous and long continued. 



Fibres. In 1910, a study of the varieties of patwa 

 (Hibiscus cannabinus) was completed at Pusa when it was 

 observed that one of the kinds, Type 3, appeared to be much 

 more suitable for cultivation than any of the others. In 



