246 



PRINCIPLES OP PLANT-TERATOLOGY. 



In a flower of Gattleya labia fa there were two sepals 

 (anterior and posterior) and two lateral petals ; no 

 labellnm. Fig. 148 shows a flower of Coelogyne cris- 

 ta ta in the same condition. 



In a flower of 0. Loddigesii var. Harrisonide, on the 

 other hand, there were two lateral sepals, and two 

 labella (anterior and posterior) (PI. L, fig. 3). 



In Narcissus biflorus every alternate whorl was 2- 

 merous (fig. 149). 



Wherever pentamery occurs in Monocotyledons we 

 see a trace of this same tendency to reduction in every 

 alternate whorl. 



A. ,t l 



Fig. 149. — Narcissus bijiorus. Dimery in the calyx, the first whorl 

 of the androecium, and the pistil, s, sepals ; p, petals ; s£ a , first 

 staminal whorl ; st b , second staminal whorl ; pi, pistil. 



F. Midler describes flowers of Hedychium which 

 were 2-merons throughout ; the usually bifid labellum 

 was in these flowers quite entire. 



Engler describes various interesting flowers of the 

 yellow rocket {Ba^barea vulgaris) which were 2-merous 

 throughout, the whorls being often so many as seven 

 or ten ; often all or most of the transverse whorls had 

 bifid leaves (including a transition to 4-mery), while 

 the median ones had simple leaves. The floral members 

 in these flowers were mostly virescent. 



Heinricher* found " incompletely 2-merous (pseudo- 

 dimerous) flowers of Iris hungarica, whose perianth was 

 constructed exactly like that of normal flowers of 

 Philydraceae. The two posterior leaves of the outer 



* The original paper has not been seen, so the facts are taken from 

 Celakovsky's account of it, and no excuse is made for quoting at length 

 from this source, the matter being of such importance. 



