CENTRIFUGAL INFLORESCENCE. 



103 



292. lu regard to the evolution of the inflorescence, that is, 

 the mode of succession in the development of the flowers, bota- 

 nists have recently observed two important distinctions, namely, 

 the centripetal and the centrifugal, the former resulting from 

 axillaiy.and the other from terminal flowers. 



293. In CENTRIPETAL iiiflorescence the evolution (blossoming) 

 of the flowers commences with those of the circumference (or 

 the base) and proceeds towards the centre (or the summit), as 

 in the Umbellifera; and the Cruciferge. 



a. The studcut will readily perceive that the ciraimfereiKe of a depressed (flat- 

 tenedanflorescence corresponds to the base of a lengthened one; and also that tlie 

 centre of the fonner answers to the summit of the latter. For when the axis, or 

 rachis, is lengthened, it is the centre which it bears along with it at its apex, leav- 

 ing the circumference at the base. 



294. In CENTRIFUGAL iiifloresccnce the blossoming com- 

 mences with the terminal and central flower, and proceeds 

 towards the lateral flowers, or those of the circumference. Ex. 

 Hyckangea, elder, and the pink tribe. 



a. ' This mode of inflorescence is generally indicated by the presence of a soli- 

 tarj- flower seated in the axils of the dichotomous (forked) branches.' All the 

 flowers are considered terminal, because they do in fact (except the first which 

 terminates the axis) terminate lateral branches successively produced at the node 

 next below the primary flower. This is beautifully illustrated in Spcrgula. 



FIG. 37. — Modes of inflorescence ; 1, centrifugal inflorescence (cyme) of Ceraslium maii- 

 mum : 2, fascicle ; 3, centripetal inflorescence (corj-mb) ; 4, spike. 



295. Sometimes we find these two modes of inflorescence 

 combinetl in the same plant. In the Composit;c, as Dr. Gray 

 remarks, the heads, which may be called the partial iutlores- 



