CH. xxviii] PHYLLODE THEORY 337 



or remaining quite rudimentary. These plants are undoubtedly 

 descended from ancestors of the normal Angiospermic type, 

 characterised by possessing a root system; but they have them- 

 selves entirely lost the ancestral capacity for producing roots. 

 Nevertheless, in both these unrelated genera, the need for an 

 absorbing organ seems to have re-asserted itself, and to have 

 been met, not by the re-establishment of root-formation, but 

 by the development of special subterranean shoots which 

 though not of the morphological nature of roots perform 

 root-like functions. This behaviour, in the case of Ceratophyllum 

 and Utrkularia^ may be interpreted to mean that a plant which 

 has entirely given up root-formation and afterwards again 

 experiences the need of roots, cannot re-acquire them, but can 

 only press some existing organ into the service, modifying it as 

 best it may. It is possible that the root-like water leaves of 

 Sahinia indicate a similar history. 



As another instance of the working of the * Law of Loss/ 

 we may take the phylogenetic history of the leaves of the 

 Alismaceae, Pontederiaceae, or Potamogetonaceae or indeed 

 of any other Monocotyledons which possess * laminae.' But 

 whether or no this illustration be accepted, depends upon the 

 standpoint adopted regarding the general morphology of the 

 leaves of Monocotyledons. The typical Monocotyledonous leaf 

 is of simple, linear to ovate form, with a sheathing base and 

 parallel veins; how is such a leaf to be compared with that of 

 a Dicotyledon, consisting, in its fullest expression, of leaf-base 

 and stipules, petiole, and net-veined lamina.'' This question has 

 naturally attracted the attention of morphologists, and an inter- 

 pretation, which has become known as the * phyllode theory ' 

 was first put forward with some reservations by de Candolle^, 

 not much less than a century ago. According to this view, the 

 typical Monocotyledonous leaf does not correspond to the 

 complete Dicotyledonous leaf, with its leaf-base and stipules, 

 petiole and lamina, but is merely the equivalent of a petiole 

 with a sheathing base. It seems to the present writer probable 



iCandolle, A. P. de(i827}. 



A. W. P. 22 



