ON METAMORPHOSIS IN PLANTS. 123 



Sparmannia africana may be cited as an example 

 derived from floral structures. The flower possesses 

 numerous stamens, of which the outer ones are modified 

 so as to form conspicuous bodies, which attract attention 

 to the blossoms, and are concerned with the secretion of 

 nectar. Accordingly, the filaments are not smooth, as is 

 usually the case, but are bright yellow in colour, and they 

 secrete, especially in sunshine, drops of fluid. These sta- 

 mens, the pollen sacs of which are either entirely absent or 

 are merely vestigial, are connected with the normal stamens 

 by a complete series of intermediate forms. Anemone 

 Pulsatilla, in which the outer stamens are also converted 

 into nectaries, similarly exhibits all the transitional stages to 

 the perfect organ. I think these facts only become intel- 

 ligible on the view that the modification of the stamen- 

 rudiment may set in at different periods, and thus the 

 formation of intermediate forms is just what would have 

 been anticipated. 



The intervening stages which, in the examples already 

 quoted, can be recognised by the unaided eye, may also be 

 definitely followed out under the microscope ; it will be 

 better, however, to consider this side of the subject in 

 connection with the root, as the conditions are simpler than 

 in the case of the leaf. 



Every one understands a " typical ' root as an organ 

 possessed of definite peculiarities of form and anatomical 

 structure. Now, these peculiarities are very intimately 

 connected with, and depend upon, the functions which the 

 root has to perform ; and, indeed, only on this view do 

 they become intelligible. There are, however, some roots 

 which, though at first behaving in all respects as normal 

 organs, subsequently cast off their calyptra, and henceforth 

 assume the mode of growth characteristic of shoots. Until 

 recently, only two cases (Neottia nidus-avis and Aut/iurium 

 longifolini) of this nature were known, but others have now 

 been traced in the case of several ferns {e.g., species of 

 Platycerium and in Asplenium esculentum). No one can 

 refuse to admit that a real metamorphosis has occurred 

 here, and one which is associated with profound anatomical 



