THE ABORTION OF OVULES 367 



of ovules and of young seeds are generally found in the constricted 

 portions, it is shown that in certain cases no such remains exist, and 

 it is argued that here we are concerned with only rudimentary ovules 

 in the flower. 



(g) Special stress is laid on the important indications given by the 

 seedless fruit, where the fruit develops under the stimulus of pollination 

 but the seeds fail. In the effects of failure of the ovules and young 

 seeds on the fruit's shape a distinction is drawn between the constant 

 failure of ovules in fruits, like those of Quercus and Cocos nucifera^ and 

 the normal failure after fertilisation of a certain proportion of the 

 ovules in many-ovuled flowers, such as those of Iris and Primula. In 

 the one case a persistent effect is produced on the fruit, in the other 

 case an inconstant one. 



(10) Much remains to be determined before one could safely 

 generalise in these matters. It will be necessary to distinguish between 

 the primal complement of ovules in a flower and the complement on 

 which the systematist bases his distinctions. Of the second all are 

 *' functionable." Of the first many may never have passed beyond 

 the primordial stage and may always exist as rudiments in the flower. 

 Thus the carpels of Anenome are described as one-ovuled, yet the beaked 

 form of the fruit has received its impress from other functionless ovules 

 which the systematist does not recognise. 



