Gkiffin. — Development of New Zealand Conifer Leaves. 59 



of the leaf, only one layer of small parenchyma being between 

 the resin-canal and the hypoderm. The elongated elements on 

 the upper surface are not nearly so long as those of the flattened 

 leaf, and are fewer in number, as we find only one row. 



The vascular bundle is like the preceding one, only very 

 much reduced, there being only three or four elements of phloem 

 and wood. The px is turned towards one of the more prominent 

 margins, as in the preceding section, and it is more obvious 

 here that the two sides nearest the resin- canal represent the 

 lower surface, whilst the two nearest the px represent the upper. 



Origin of Flattened Form. 



Now, it has already been pointed out that from the order 

 of succession and the arrangement on the stems the awl-shaped 

 leaves should be considered the more primitive. The first leaves 

 are formed while the cotyledons are still inside the endosperm, 

 and hence are shut up between them. These young leaves 

 have therefore a very constant environment in the successive 

 generations. The leaves, however, after the cotyledons have 

 expanded are subjected to much more varying conditions, and 

 hence some slight variations in form might prove advantageous 

 under a given condition, and thus, in course of time, become 

 " selected." In this case it would seem probable that the 

 voting plant at a certain period of its history found that, after 

 the store of food had been used, the greatly reduced awl-shaped 

 leaves presented an inadequate surface for assimilation. Hence 

 by natural selection it may have gradually acquired the more 

 flattened form, which now appears at a very early stage in the 

 cycle of development. This theory is borne out by a com- 

 parison of the transverse sections of the two forms, where we 

 find out also the detailed evidence of the change. It was seen 

 that in the awl-shaped leaves the elongated elements were 

 absent on the morphologically lower surface of the leaf, and 

 only one row was present on the upper. In the flattened form, 

 however, we find elongated elements on both sides of the bundle, 

 and these are also longer and more numerous on the upper 

 surface of the bundle. The leaf has not actually flattened, in 

 the sense of detracting from the thickness to add to the width, 

 but has extended itself out on two sides by the elongation of its 

 parenchyma. By this extension a flattened form of leaf has 

 arisen, for the width of the new leaf is much greater in proportion 

 to its thickness. We may therefore speak of the extension as 

 a flattening process — i.e., the leaf has become flattened in the 

 median plane. 



The flattening, further, has taken place in such a direction 

 that a dorsi- ventral arrangement of the leaves, in two rows, 



