IN CEYLON. 85 



D. Ebenum is due to the interpolation of a female flower in a 

 male cyme. The female flower when thus occurring 

 presents no differences from the female flower of dioecious 

 trees, except in that it is pedicellate. In no case have 

 I met with the monoecious form due to the replacement of 

 a female flower by a male. 



The occurrence of true male and female flowers in the same 

 inflorescence is of importance, since the characters of the 

 accessory and essential whorls in the male and female are 

 often so widely different. It has been suggested by many 

 observers that since the flowers, the habit, and foliage of what 

 are regarded as the male and female trees of the same species 

 often show great differences, we might be erroneously 

 considering male and female trees of what are really separate 

 species as being of the same species. Subsequent observa- 

 tions demonstrated the frequent occurrence of polygamy, and 

 for a time these gave support to the above contention. Now, 

 however, we have seen that the true male and true female 

 flowers possessing different types of accessory and essential 

 whorls can occur in the same inflorescence, and we can 

 therefore adhere to the former ideas as to the true dioecious 

 form of some species. 



Characters of Accessory and Essential Whorls. 



Calyx. The calyx is always inferior and gamosepalous ; 

 sometimes glabrous or nearly so, as in D. sylvatica, at other 

 times densely pubescent, as in D. Ijlmbryopteris and D. 

 pruriens. 



In the male flowers the calyx segments are relatively 

 small and accrescent, whereas in the female and hermaphro- 

 dite flowers they attain considerable dimensions and the 

 segments project outwards. 



In the Ceylon species the calyx segments are usually very 

 distinct, varying in number from two to seven ; the common 

 number of segments is four or five. In some species, how- 

 ever, the calyx segments cannot be distinguished, as in D. 



