OF INFLORESCENCE. 17 



analogous to the discs which support the partial flowers 

 of the Umbel] iferae. 



In all the inflorescences which we have enumerated the 

 law of expansion is simple and uniform ; everywhere the 

 lower and external flowers open first ; and the flowering 

 proceeds consequently from the base upwards in the 

 spike and raceme, and from without inwards in the 

 corymbiform racemes and the umbel ; it proceeds from 

 the base upwards in the spiciform or elongated capitula, 

 and from without inwards in the flat ones. This regular 

 progress of the expansion Rceper has called by the name 

 of Centripetal. It must, however, be remarked that, 

 as regards compound spikes or racemes, the central axis, 

 which is the prolongation of the stem or principal 

 branch, flowers before the lateral ones, each of which 

 follows in its turn its development in the same direction. 

 The only exception which I know to this mode of 

 development, is that which certain Dipsaceae present, the 

 flowering of which often commences in the middle of 

 the spike : this anomaly must be connected with some 

 peculiarity in the vegetation of these plants ; for, as to 

 their form, they cannot be separated from centripetal 

 inflorescences, and this mode is presented regularly in 

 the other species of the family. 



Some of the inflorescences which have been enume- 

 rated in this section may be combined together. Thus, 

 the flowers of the Gramineae are, as we have said, dis- 

 posed in small distichous spikes, which have been called 

 spikelets ; and these, more or less pedicellate, are dis- 

 posed in panicles, sometimes fvery lax, at others 

 more or less compact ; the flowers of Carex are dis- 

 posed in close spikes, which are arranged in a raceme 

 along the central axis ; those of Papyrus are dis- 

 posed in spikelets/Avhich are pedunculate and arranged 

 in an umbel at the top of the stem ; the flowers in 



VOL. II. C 



