Geiffin. — Development of Neiv Zealand Conifer Leaves. 63 



starting to be formed on all sides of the bundle, seemingly aris- 

 ing directly from the px. This is a later development, arising 

 out of the increase in parenchyma tissue, for there is not nearly 

 so marked a development seen in the awl- shaped leaf of the 

 same stage. In the mature leaf we see this development car- 

 ried further, and transfusion tracheids occur on all sides of the 

 bundle, and arising in some cases from the px on the ventral 

 surface. This leaf would form a strong support for Mr. Wors- 

 dell's theory, unless the intermediate forms had been studied. 

 We may regard here the transfusion tracheids on the ventral 

 surface as a later development of centripetal xylem, arising on 

 account of the needs of the leaf, but not as modified primitive 

 centripetal xylem. 



We will now pass to two species of a different genera — 

 Dacrydium cupressinum and D. Kirkii — and show where they 

 differ from the species of the preceding genus. We will take 

 D. cupressinum first, as it shows in its foliage many points of 

 resemblance with the last species. 



Dacrydium cupressinum (Eimu). 



Of this species I was fortunate in finding all forms growing 

 under the same conditions, from young germinating plants to 

 mature foliage. The mature leaves of this species are very 

 hard to distinguish from those of the kahikatea, especially when 

 separate from the mother tree. Both are awl-shaped, and 

 arranged spirally, closely adpressed to the branches. The leaves 

 of the rimu are, however, slightly longer, and not quite so closely 

 adpressed to the stem as those of kahikatea. Both trees, when 

 growing amongst other trees in the forest, lose their lower 

 branches. The height of the tree thus makes it very hard to 

 distinguish the difference in foliage when viewed from the ground ; 

 but these trees can readily be distinguished by other points. 

 One of the most important of these is that, while the lateral 

 branches of the kahikatea are erect, those of the rimu are pendu- 

 lous. Hence the rimu is greatly used for ornamental purposes, 

 while the kahikatea is but rarely so used. If grown in the open, 

 as in cultivation, the rimu may grow to a great height while 

 still keeping pendulous branches low down on the main stem. 

 An analogy to this was seen in the kahikatea. In the young 

 stages, however, there is a great difference in the appearance 

 of the young plants of these two species : this is due to the 

 absence of dimorphic foliage in the rimu. Here we find only 

 narrow awl-shaped leaves arranged spirally round the stem. We 

 find little or no flattening of the leaves, though there is a slight 

 tendency in the earbest stages to flatten each side of the bundle, 



