182 JOURNAL OF THE ROYAL HORTICULTURAL SOCIETY. 



ray florets play a leading part in enhancing the " attractiveness " 

 of the capitulum, and indeed in some genera play no other, for 

 they are sometimes void of both sexual organs, and are barren or 

 neuter. The florets of each capitulum open, as has been pointed 

 out, first on the circumference, and so proceed inwards, and the 

 same order is observed in the development of the sexual organs 

 of the flower, the rule being that the male organ is ready first, 

 roughly speaking some " rings " ahead of the female, so that it 

 happens that a certain stigma is not likely to receive the pollen 

 from an adjacent stamen. And Nature makes another provision 

 to the desired end of impeding self- fertilisation, and we see this 

 in the history of the development of the floret. The filaments of 

 the stamens are contractile. At the outset the united anthers 

 are not only conjoined, but themselves enclose the stigma. At 

 this time the two lobes of the stigma are not separated, but are 

 pressed closely together, the true stigma surfaces, which are 

 the inner adjacent faces of these lobes, being completely protected 

 from the pollen. When at length the opening of the anthers 

 sets the pollen free it is discharged into the vacant space between 

 the anther lobes and above the stigma. The style then lengthens, 

 and the stigma, with its closed and abutted faces, pushes up and 

 brushes out the pollen from the tube of the anthers. At length, 

 when the stigma has thus swept away the pollen of the particular 

 floret to which it belongs, its two lobes separate, and it is at 

 last in a position to receive pollen ; but this must be the pollen 

 from another floret, and possibly from another capitulum 

 altogether. 



Such is the progression of development of the florets regularly 

 from circumference to centre. 



The practical lesson to be derived from the foregoing is this, 

 that, in the absence of insect agency, self- fertilisation is almost 

 an unknown quantity in the problem, and the matter of selection 

 of the parentage of the coming offspring is very much in our own 

 hands. 



Furthermore, we obtain from watching the regular develop- 

 ment of the rings of pollen-bearing stamens a very useful 

 indication as to where we are likely to find the stigmas in a con- 

 dition to receive the pollen which we wish to convey to the ovary. 

 In practice it will be found a sufficient guide to take it to be a 

 rule that when the stamens in the inner section of the disc are 



