20 Mr. T. Smith on certain Sj)ecies of 



corolla. These differences are less obvious after the flowering 

 is past ; for, the stamina being retracted, the hermaphrodite is 

 much more like the female : as to numbers, the hermaphrodite 

 is the most prevalent. 



In Serratula tinctoria, in which Mr. Bro^vn first pointed out 

 the existence of female plants, I have not been able to satisfy 

 myself that males are to be met with ; for in the antheriferous 

 plants I have always found the ovaria impregnated. The seeds 

 of the female differ in being larger than those of the heniiaphro- 

 dite. In this species plants occurred which showed a regular 

 gi-adation from the female to the hermaphrodite ; in one, the 

 antherae were much smaller, shorter, and more imperfect than 

 they most frequently are found in the female : in another they 

 were as much larger, projecting, and embracing the style as in 

 the hermaphrodite, but containing only a few grains of abortive 

 pollen. 



The numbers of the female and hermaphrodite are nearly 

 equal. The stigma of the female is developed very soon after 

 the flower opens; in the hermaphrodite, on the contrary, it 

 does not appear until the pollen of its own antherae is dispersed, 

 the style remaining undivided to the apex till this period ; the 

 aid of the antherae of some adjoining flower consequently be- 

 comes necessary for the purposes of impregnation. 



This is a striking example of a mode of impregnation which, 

 according to M. Cassini, prevails nearly throughout the whole 

 family of the CompositcE, and which renders the presence of 

 two flowers at the very least necessary to the impregnation of 

 either; constituting, in fact, a sjiecies of monoecious inflore- 

 scence ; and as it requires some external aid for its completion, 

 forms a transition to the decided separation of the sexes in di- 

 stinct florets, which are further removed into distinct capitula 

 in the monoecious genera Xanthium and Ambrosia, and still 

 further in the dioecious plants. 



This process is analogous to that which takes place in a few 

 instances in the animal kingdom, in what are on this account 

 termed androgynous animals, of which the Helix hortensis is a 

 well known example. In the androgynous animal, although 

 it has both the male and female oi-gans complete, the one can- 

 not be impregnated by the other on account of their relative 

 position : in the f ndrogynous flower, impregnation is pre- 

 vented by the organs of the two sexes not being developed at 

 the same tune. 



I am not aware that any particular term has been adopted 

 to designate flowers of this kind ; but as they are not confined 

 to the family of the Composite, it might be useful to point them 

 out by an appropriate name, and androgynous seems strictly 



applicable. 



