950 JOURNAL OF THE ROYAL HORTICULTURAL SOCIETY. 



one of the latter was deeply forked. In the next case to be considered, 

 however, the virescence must have set in at an earlier period, viz. before the 

 maturity of the corolla, which, as seen in figs. 216 and 210, h, has developed, 

 in the majority of florets, to an abnormal size, at the same time becoming 

 perfectly green instead of yellow. The non-development of the stamens 

 and pistil in these cases is doubtless correlated with this exceptional growth 

 of the corolla. 



In some capitula the disc-florets, besides becoming virescent, had pro- 

 liferated into shoots (figs. 212 and 210, h) ; the virescent corolla itself was 



Fig 215. — Capitclum showing both Fig. 216. — Capitulum with great Develop- 

 Ligclate Corollas and Virescent ment of Virescent Corollas of Disc- 



Carpels of a few Disc-florets. florets. 



carried up on a long stalk, and the axis of the shoot was continued for a 

 considerable distance beyond it, bearing, not far from its summit, a pair of 

 small foliage-leaves : the exceedingly vegetative and luxuriant carpels ; at 

 the top of all was a young secondary capitulum or flower-head (fig. 210, h). 

 In some of the more peripheral florets an early stage of the proliferating 

 axis could be observed as a minute shoot springing from the centre of the 

 floret, bearing 1-3 small green leaves, probably the carpels, which protrude 

 beyond the mouth of the corolla ; between them and borne at the tip of 

 the shoot was a tiny capitulum. The earliest stage of this proliferation, 

 where the shoot is as yet confined within the ovary — in other words, 



