TO BOTANY. 41 



bat thaMlie contrary is the very characteristic of the 

 radiate ; now this being attended to, the manner of 

 the impletion will be easily understood. Compound 

 flowers gain their- impletion two ways, either by the 

 radius or the disk. We shall begin with the first. 



Impletion by the Radius is when, by the multipli- 

 cation of the radius, the disk of the flower is filled 

 up; as in Helianthus, Calendula, Chrysanthemum, 

 Anthemis, Matricaria, Ptannica, Tagetes, find the 

 species of Centaurea, called Cyan us. In this sort of 

 impletion, which belongs only to radiate flowers, it is 

 observable that all the florets which fill up the disk 

 follow the conditions of those of the radius ; so that 

 if the florets of the radius in the natural flower have 

 a pistillum, all those of the full flower will have one 

 also, as in Matricaria, Bellis, Chrysanthemum, and 

 Tagetes ; or if they haye no pistillum, then it will also 

 be wanting in the full one, as in Helianthus, Calen- 

 dula, and Centaurea; and the same holds true of the 

 male part also ; for as the florets of the radius in the 

 natural flower are never furnished with anthera?, so 

 these are wanting also in those of the full ones. This 

 last remark is of great use to distinguish a radiate full 

 flower, rrom a iigulate natural one ; which might be 

 confounded in many cases, were we not apprised 

 that there are antheroe in the latter, but none in the 

 former ; by this rule, in Chrysanthemum, Helian- 

 thus, Calendula, and Tagetes, when the disk is de- 

 stroyed by the multiplication of the radius, we know 

 by the defect of antherae, that it is only the luxu- 

 riancy of a radiate flower, as in Hieracium, Leontodon 

 and Sonchus ; by the presence of the antherae we 

 know the flowers to be Iigulate and natural. 



Impletion by the disk is, when there is no multi- 

 plication of the radius : but the corolliilos of the disk 

 run out into length, and have their brims less divid- 

 ed : this manner of impletion seems to concern only 



