ON METAMORPHOSIS IN PLANTS. 125 



first leaves, as is usual, do not twine, but in the succeeding 

 ones the stalks of the leaflets gradually elongate, while 

 the blade-portion concomitantly suffers reduction until typical 

 tendrils are produced, and the same process obtains in many 

 other plants. To quote one more example, the tendrils to 

 Cobcea scandens owe their existence to luxuriant branching of 

 the terminal portion of the leaf; the hooks of the tendril 

 represent very much modified leaflets, while the arms are 

 derived from the stalks of the pinnae. Occasionally one 

 meets also with transitional forms which are intermediate 

 between tendrils and leaflets, and which one can easily 

 recognise as such without the aid of the microscope. In 

 other cases, the entire leaf rudiment (Blattanlage), and not 

 merely the leaf stalk, may develop into a tendril. Thus in 

 the first-formed tendrils of Beuincasa cerifera, a cucurbi- 

 taceous plant, it is easy to recognise the rudiments of a leaf- 

 blade, but this becomes increasingly difficult in the tendrils 

 which are formed later on. 



As regards the Phyllodes of the Australian Acacias, they 

 are often spoken of as flattened petioles, the leaf-blades of 

 which have completely vanished. But, after what has been 

 already said, it is obvious that this view is untenable, and as 

 a matter of fact, the blade of the leaf is clearly enough 

 visible, only it remains very small. This explains how it 

 may happen that in some species the phyllodes are accom- 

 panied by fully developed foliage leaves, and observations on 

 the seedling plants also indicate plainly the steps by which 

 a real metamorphosis has -been effected. Additional ex- 

 amples might easily be quoted, but they would only illustrate 

 the same point, namely, that metamorphosis is an actual 

 objective fact ; that the rudiment which is normally destined 

 to develop into a definite organ may change its ordinary 

 proportions during development, and thus give rise to 

 another organ, and that this change of form is constantly 

 associated with an alteration of function. Thus we have 

 come to limit the conception of metamorphosis, so that it 

 only applies to the results produced by the process of division 

 of labour amongst homologous members. We recognise, in 

 dealing with the homologous members of a branch system, 



