192 Journal of the Asiatic Society of Bengal. [May, 1913.] 
The flowers collected and examined by me revealed a great 
deal of variation which is recorded at the end of this paper ina 
tabular form. It will be noticed that altogether 457 flowers 
were examined. Thenumber of stamens always agreed with 
that of the petals in the same flower. The sepals agreed in 
number in 80% of the total with the petals and stamens, showed 
a deficiency in 5%, and an excess in 15% of the flowers 
examined. The variation in the number of parts in all the three 
whorls lay between 5 and 8. Fifty-one per cent of the flowers 
showed six parts in the petals and stamens (but not in the calyx 
at the same time) and 38% presented six parts in all the three 
outer whorls in the same flower. The pistil was trimorphie while 
the stamens were only dimorphic. Of every hundred flowers 
51 had long stamens and 49 short ones. Regarding the pistil, 
differences but also the trimorphism of the pistil, we get 
altogether fifty-five different forms of the flower. I am not 
aware of this record exceeded anywhere. 
