THE ANGIOSPERMAE 



1131 



indeterminate and that it only becomes, strictly speaking, a primordium or 

 rudiment when the course of its future development has been determined. 

 The flower may thus be thought of as divided into a series of horizontal, 

 biochemical zones, the lowest of which is hormonically sterilized as sepals. 

 The balance between the zones of maleness and femaleness is closely 

 adjusted, which may be attributed to the clear-cut opposition between the 

 controlling influences, but the sterility control appears to spread upwards 

 from the calyx into the adjacent male zone, leading to the sterilization of a 

 certain number of its members as petals. This balance is not exact and some 

 hesitancy or indefiniteness may appear at the line of junction, with the 

 production of intermediate organs as petaloid stamens of varying degrees of 

 sterility. There is thus no positive correlation between the numbers of petals 

 and those of stamens. In primitive flowers with numerous free parts, 

 such as the Raniincuhis type, the two sets of parts are indeed vicarious, that 

 is to say they are negatively correlated, an increase in one set corresponding 

 to a decrease in the other, as would be expected if the petals are transformed 

 stamens. The orderliness of the flower may therefore be due to hormonic 

 control rather than to the geometrical requirements of close packing on the 

 receptacle, which does not in fact begin to operate until the primordia are 

 partially developed and their nature already determined. 



The question whether the immediate ancestors of the Angiosperms 

 were hermaphrodite or unisexual has been for long a subject of research 

 and speculation, since on the answer depends what view we take of the 

 evolution of the group. In the first case the receptacle produces both types 

 of sporangial appendages and in the latter case it produces only one, that is 

 to say that one of the reproductive zones is missing. If we adopt the com- 

 parative method we can find facts which appear to support either theory. 

 The most widely held view has been that a hermaphrodite type with 

 numerous floral parts, like that of the Magnoliaceae, presents the most 

 primitive mode of organization among living Angiosperms, and that the 

 unisexual condition has been evolved from this by a process of reduction. 

 This fitted in well with the theory that the angiospermic flower was evolved 

 from large bisporangiate strobili of the kind represented by the fossil 

 Bennettitales (Fig. 1105). 



Many stages intermediate between the hermaphrodite and the unisexual 

 state can be found among living flowers. A series may be traced leading 

 from simple sterility of either androecium or gynoecium, through regressive 

 degrees of imperfect development, towards the complete disappearance of 

 one or the other. It is inconceivable that such a series of intermediate 

 stages can be read otherwise than as a process of elimination of one or other 

 of the sets of sexual parts. The androecium is the more frequently affected 

 but abortive gynoecia are also known. Even in the multipartite Ranuncula- 

 cean flower, Caltha palustris, the vascular anatomy shows that the apex of 

 the gynoecium has disappeared and that in the double variety grown in 

 gardens it has vanished altogether. 



As we shall see later in connection with pollination it is not uncommon 



