THE FLOWER. 



35 



neighbors, would result in a compound five-celled ovary. Suppose them 

 united to each other from base to apex, and one common stigma might do 

 for* all ; let union take place only half way up, and there would be a com- 

 pound ovary, but with five styles and five stigmas. The pistils of fetid 

 hellebore (Fig. 63) are united at the base only, and are considered simple 

 by some, compound by others. 



Of the pollen and ovules little need be said here, save that they have a 

 strict analogy with the fertilizing element and the ovule of animal repro- 

 duction. Both exhibit a great variety of forms. 



Fig. ()3.— Pistils of fetid hellebore 

 united at the base. 



Pig. C4. — A lily. Parts of the flower in threes (3-merorts). 



Having thus briefly considered the organs which compose the flower, 

 we will now examine their numerical disposition and their arrangement 

 upon the receptacle. 



In exogenous plants the parts of the flowers are commonly in fives or 

 fours, or in multiples of those numbers ; and however much they may 

 differ in this respect, they are never completely in threes. In endogenous 

 plants, on the contrary, the habitual arrangement is in threes. This 

 should be borne in mind, since it is another prominent and characteristic 

 mark of difference between these two great divisions of flowering plants. 



In speaking of a flower in respect to the numerical arrangement of its 

 parts, it is said to be 3-merous, 4-merous, or 5-merous (Figs. 64, 

 65, and GQ). 



But here, as elsewhere in plant life, there is endless diversity. A strictly 

 5-merous flower should have five sepals, five petals, five stamens, and five 

 pistils, or multiples of this number. Now, in reality such a flower is rare. 

 It would be much easier to find one with five sepals, five petals, ten sta- 



