200 STUDIES IN THE POLLINATION ON INDIAN CRofS 



These results prove that in this species as in many others of this order, 

 autogamy does not occur unless the stigmatic surface is first stimulated. One 

 consequence, therefore, of insect visitation is self-pollination. Stimulation 

 of the stigmatic surface does" net appear necessary however for cross-pcllina- 

 tion. In 1916, a number of flowers were artificially crossed without first ( f all 

 rubbing the stigma. All formed pods and set seed. These results indicate 

 that cross-pollination by means of insects is the first object and that only if 

 this fails is seed formation provided for by self-pollination. 



Natural cross-fertilization. All the facts point to the possibility of exten- 

 sive natural crossing in this crop. That this is so was confirmed by an experi- 

 ment carried out at Pusa with two cultures cf very different habit which were 

 growTi side by side and allowed to flower freely for some years. The morpho- 

 logical differences of these cultures are summed up in the table : — 



Local Pusa variety Jubbulpore variety from the Central 



' Provinces 



1. Seed. Small, shiny, black. Large, dull black or greyish. 



2. Seedlings. Germination slow, small. Germination rapid. Seedlings large with 



with reddish downy cotyledonary green glabrous cotyledonary leaves and 



leaves and stems. stems. 



3. Plants. Short, late maturing, with many Tall, early maturing, with a few short sparsely 



flowered spreading branches beginning flowered parallel branches beginning at 



at a jjoint about four feet from the a point about eight feet from the ground, 



ground. 



At first, the natural differences between these two kinds were maintained 

 but they somewhat rapidly disappeared and in two or three years the cultures 

 were indistinguishable. Crossing had evidently taken place in all directions 

 and eliminated the well marked differences in habit. 



This result led to a detailed examination of the local Bihar crop in order 

 to determine its constitution. The greatest amount of variaticn was noted 

 in the standard of which twelve distinct types could be distinguished, differing 

 in the colouration of the veins, in the depth of yellow and in the amount and 

 distribution of a reddish tinge existing with the yellow. Several colcur factors 

 therefore occur in the flower and there is little doubt that the crop contains 

 numerous forms differing very slightly from one another. 



Improvement. The fact that no setting takes place if the flowers are pro- 



-tected, that seed formation is dependent en insect visits and that extensive 



natural crossing takes place render the improvement cf this crop a difficult 



matter. Variety trials, extending over several years, are only possible by 



