166 



vation, which begins .very early to be somewhat emarginated 

 above. When a Httle older this upper surface assumes the form 

 of a disk inclined greatly towards the bract, exhibiting in front 

 two distinct teeth, with a furrow below them. By degrees, that 

 part of the stigma where the furrow is, begins to run up, more 

 like the other side. In the full-grown condition, the teeth, 

 which originally occupied the summit of the style, are found as 

 a pair of little papillae on its back, and the said furrow becomes 

 the real stigma, in which are ultimately found the capillary 

 cells between which the pollen-tubes make their way into the 

 ovary. 



The most interesting fact in the history of development 

 just given, appears to me to be the formation of the anther 

 and labellum in this flower. The anther and labellum, as we 

 have seen, do not extricate themselves from a mass remain- 

 ing in the centre of the flower, as in the other flowers observed, 

 but a number of furrows, corresponding to the arrangement of 

 the inmost circle, are formed in the inner side of a circle of 

 organs already developed ; not divided, it is true, but suffi- 

 ciently characterized by form and position. A little later we see 

 the anther and labellum forming one connected body, and the 

 lobes of the latter then become apparent in this. Here there- 

 fore we might assume the production of a circle of organs by the 

 radical deduplication of the coroUine organs, which would be 

 borne out by the development. But leaving out of the ques- 

 tion that in Calathea, for instance, where the necessity of such 

 a circle of organs exists quite as much, there is not a similar 

 mode of development, it might equally be supposed here, that 

 the two circles were originally confluent into one mass. And it 

 may be urged against the first hypothesis, that at the time when 

 the anther and labellum are still connected, the lobes of the latter 

 are already indicated, at all events three of them, the middle one 

 of which, alternating with the corolla, is unfavourable to this 

 theory. But this early division of the labellum (so contrary to 

 what is observed in other foliaceous organs) is that which, in 

 my opinion, may on the other hand be brought to support the 

 assumption that the original organs lying between the six outer 

 organs of the flower and the pistil are to be regarded as stamens. 

 Thus the course of development here is as favourable as possible 



