452 Proceedings of the Royal Irish Academy. 



walls of tlie passage are formed by small elongated cells, which in 

 transverse section present a rhomboidal outline. The wall of these cells, 

 which is next the passage, is strongly cuticularized. The cells adjoin- 

 ing the other side of these lining cells are usually strongly thickened. 

 As we trace this passage downwards in the leaf we find that it opens out 

 towards the base of the leaf into a more or less triradiate appearance in 

 transverse sections, and further down it becomes nearly circular (fig. 1 2). 

 The similarity of the structure and arrangement of the cells lining this 

 passage to those of the epidermis on the outside of the cylindrical leaf 

 suggest that this passage is in reality part of the epidermis cut off in 

 the interior of the leaf. This surmise is shown to be correct by tracing 

 the passage upperwards in the leaf. The passage throughout its entire 

 course runs almost axially in the leaf. About 3 mm. from the pointed 

 apex of leaf it gradually approaches the surface of the leaf opposite to 

 the large odd bundle of the central five bundles which are continued 

 up to this distance in the leaf. Finally, at a position very close to the 

 apex, it reaches the surface (fig. 21). At this point the bundles of the 

 leaf are reduced to three. Longitudinal sections through the base 

 of the foliage leaf and the apex of the stem which bears it, shows the 

 passage opening out into a funnel-shaped chamber, which covers the 

 apical bud. This latter, in all cases examined, had failed to develop. 



In order to learn how this peculiar passage in the leaf was deve- 

 loped transverse and longitudinal sections of the axillary buds were 

 made, and after many trials I succeeded in obtaining some sections of 

 buds just at the moment of forming the foliage leaf (fig. 19). Prom 

 these it could be seen that the apex after the formation of the foliage 

 leaf is depressed and that the leaf itself arises as a collar of tissue round 

 the apex, which grows up round the latter. Simultaneous growth in 

 thickness and in height of the collar over the apex forms a narrow 

 channel leading down the axis of the leaf to the apex. 



From the anatomy, then, and the development of this peculiar 

 leaf of D. teretifolium we may conclude that this passage running 

 down the leaf represents morphologically the upper surface of the 

 leaf, which we may suppose was derived from a cylindrical leaf like that 

 of V. teres by a continued infolding of the upper surface, which in the 

 latter orchid is represented by a shallow groove. As this groove got 

 deeper the sides fused across, till finally what was originally the upper 

 surface becomes completely cut off from the exterior, except for the 

 small opening at the apex of the leaf. 



In this connexion it is peculiarly interesting to find that in 

 one orchid, at least, the leaves are permanently in an intermediate 



