i82 NATURAL SCIENCE. Sept.. 



— the sJiieU — lying flat over the surface of the leaf. This shield has a 

 central group of cells, and a thin extended margin composed of cells 

 or cell-walls. Underneath the base is a group of three cells on top of 

 one another ; they are protoplasmic and have thin uncuticularised walls 

 (except sometimes the lowest wall is thick but pitted). These three 

 cells are " passage-cells." 



If a drop of water be placed on a leaf of Tillandsia usneoides, the 

 silvery-grey appearance is locally replaced by a green colour. 

 Examination with a microscope reveals the fact that at first the 

 central cells of the shield contained air, but on supplying water these 

 cells become filled with it. Placing a drop of caustic potash or strong 

 sodium chloride solution on the leaf, it can be seen that the colour 

 changes or plasmolytic phenomena commence first in the epidermal 

 cells nearest the scaly hairs, and gradually advance to more distant 

 cells. Finally, the cells composing the sheets have their lower walls 

 thin and uncuticularised, as also are the walls of the stalk and passage- 

 cells ; whereas the rest of the epidermal cells have thick cuticularised 

 walls. Again, terrestrial bromeliads which have absorptive roots 

 possess £caly hairs which are incapable of being wetted. These facts 

 prove that the peculiar scaly hairs of the epiphytic bromeliads absorb 

 nutritive solutions. The upper wall of the scaly hair is thick and heavily 

 cuticularised. It functions as a mechanism to prevent evaporation 

 of water. Accordingly when those hairs are sunken it is relatively 

 thin, but in protruding hairs it is thick. When the central cells of the 

 shield contain air their lateral walls are crumpled so that the shield 

 lies close above the surface of the leaf (thus preventing loss of water 

 by evaporation) ; but when these cells absorb water, their walls 

 straighten and thus raise the shield above the general surface. 



It is impossible to mention all Schimper's interesting observations 

 on this part of the subject. However, he points out that some terres- 

 trial bromeliads possess absorptive scale-hairs, and he makes out a 

 good case for the view that the possession of these scaly hairs in 

 epiphytic Bromeliaceae is not adaptive, but that these plants become 

 epiphytic because they already possessed these curious hairs. 



General Conclusions. 



It has been pointed out that in the first three groups forming 

 Class I. the nutritive solution is absorbed by the roots. In the first 

 group no provisions are made for the collection of water or humus 

 otherwise than from the surface of the host-plant. This group com- 

 mences with the simplest forms which are entirely terrestrial in 

 structure, and only grow on the lower parts of trees in damp and ill- 

 hghted spots ; but the rest of the group are characterised by a great 

 development of the surface of the root- and leaf-systems. The leaves 

 have a thick cuticle, etc., to avoid excessive transpiration : and internal 

 water-reservoirs are usually present. But owing to the poverty of the 

 supply of water and humus, and to the simple measures adopted in 



