290 Journal of the Asiatic Society of Bengal. [N.S., XIII, 



Nomia oxybeloides visits flowers chiefly for gathering 

 pollen, and if the whole of the pollen from a flower could be 

 gathered by the bee in one visit, its needs will have been satis- 

 fied when a few flowers have been visited. In nature, however, 

 the flower throws out the greater part of the pollen when visited 

 by the bee, and the insect can gather only the few pollen grains 

 that may still be left on the exploded stamens. The bee has 

 therefore of necessity to visit a very large number of flowers, be- 

 fore it can get a sufficiently full load. 



When the flower explodes, the bright red wings, whicli 

 form the most conspicuous part of the flower, are carried away 

 by the interlocked keel to a different position and makes it 

 very inconspicuous ; and in its place the pale staminal column 

 is made to appear. This helps the insect to go straight to 

 unexploded flowers. The arrangement saves the time of the 

 busy insects, and is a case of distinct help rendered by the plant 

 to its insect-visitor. 



When Nomia oxybeloides visits flowers, it usuallv visits 

 only one or two flowers in one individual plant, and passes on 

 to some other, even if there happen to be many unexploded 

 flowers on it ; and may visit it again a second or a third time. 

 1 his may appear to an observer a rather strange behaviour ; 

 but we know that it is just what the plant requires, Xenogamy 

 being thereby brought about and Geitonogamy prevented. 



JNow we will pass on to the very interesting plant A. vagi- 

 nalis which belongs to the second group above referred to. 

 ine floral mechanism agrees in all its details with the last spe- 

 cies, A rugosus, except in some of the points mentioned already. 



Ibis trailing plant has a gregarious habit and also branches 

 copiously; so much so that it completely covers the ground 

 v.berever it is found growing. The branches of neighbouring 

 plants get intricately interlaced, and in one case on an exam- 

 ination of a circular patch of ground 7 inches in diameter, I 

 was able to trace branches of 9 separate plants. Each branch 

 naci a number of flowering branches. Three separate count- 

 ings oi inflorescences in one square foot of ground gave me an 

 average of about 90 inflorescences, and one inflorescence there- 

 tore occupxes only 1-6 square inches of ground on an average. 



another ^ ' great ma J orit y either touch or cross one 



™™k S *¥ P ollination ^ork has to be hurried through a large 

 2 er of open flowers, I thought I could find out the insect- 

 visitor ol this plant very easily ; but I was quite disappointed. 

 Although numbers of exploded flowers could be noticed, I could 

 J?,,T, out the msect > that must have visited the flowers and 

 caused them to explode. I was very much annoyed at mv foil- 



NhSn* 7 * f mVi \ at detectin g the 'insect, and with the resolve 



1 rbr! i i mg dmvn the insect that had been dodging me, 

 uirectea my attention to one particular inflorescence ; and 



