63S MR JOHN M'LEAN THOMPSON ON 



basal flanges nor grooves were present. Despite the fact that the pinnae varied 

 greatly in size, their venation was invariably a sympodium. This leaf was inserted 

 upon the upper surface of the rhizome, and might be regarded as a pinnate leaf which 

 had suffered extensive reduction of its rachis. 



The second step in leaf reduction is shown in fig. 9. It represents a leaf which 

 measured roughly 1 inch in length. The numerous pinnae were crowded upon a 

 condensed rachis which measured little more than - 5 inch in length. The venation 

 of the majority of the pinnae was a simple midrib, but no vascular supply was dis- 

 covered in several of the smaller terminal pinnae. At the base of the rachis an 

 abortive branch arose as a short and slender projection. The nature of this branch- 

 ing will be discussed later. In this leaf, then, is seen a further reduction of the 

 rachis, and a simplification, and even total loss, of the venation. 



A third step in this reduction process was shown by a leaf which is depicted in 

 fig. 2. Of the | inch which it measured, the greater part consisted of petiole 

 of the smaller type of filiform leaf already described. The pinnate portion of this 

 leaf had been damaged, but the vascular supply to each of the four persisting pinnae 

 was a simple unbranched vein. As will be shown later, when the anatomy of these 

 reduced leaves is being described, the rachis of this leaf had been very short, and 

 the number of pinnae which had been destroyed was probably very small. There is 

 here, in fact, a .leaf in which the rachis has been reduced to a small and condensed 

 column, bearing a few small pinnae with the simplest possible venation. 



The fourth step in leaf reduction remains to be considered. It is illustrated 

 by a small leaf, g inch long, which was found among the hairs on the side of the 

 rhizome {fig. 12). Its tip was hooked, suggestive of an original circinate vernation, 

 and it bore almost terminally two small pinnae, to which the vascular supplies 

 were simple unbranched veins. There is here the minimum of pinnation, and 

 one step further in reduction would lead to the smaller type of filiform leaf. It 

 is a far cry from the elaborate pinnate leaf to this smaller type of filiform leaf, 

 but the reduction series here briefly outlined can bridge the gap between them. 



It may be asked w T hat is the position of the larger filiform leaves in this varied 

 aggregation of leaf types. It is my purpose to show in succeeding paragraphs 

 that the larger and smaller filiform leaves alike are reductions from, or replace- 

 ments of, larger pinnate types. It is believed, indeed, that the leaves of Platyzoma 

 have undergone drastic morphological changes," resultant from the present rhizo- 

 matous habit and extreme xerophytic conditions. But the small filiform leaves, 

 and the pinnate leaves showing reduction, were rare occurrences in my specimen, 

 while, on the other hand, the larger filiform leaves were numerous. I am of the 

 opinion that, when abundant material has been examined, it will be found that 

 the larger filiform leaf is the typical reduction leaf of the present day, just as 

 the unbranched pinnate leaf, and not the branched pinnate leaf, is typical of 

 Platyzoma in its present condition. The forked pinnate leaf, the reduced simply- 



