90 ORIGIN OF STELE [ch. 



was not continuous or progressive. A certain small set of 

 proeambial elements were converted together and then the 

 process ended. There is not. the slightest sign of a leaf trace 

 origin of the cauline stele. In fact we know from Psilophyton 

 that the conversion of the proeambial elements took place in the 

 main axis and in the branches quite independently (probably at 

 different periods) and further that these foci of lignification were 

 not in continuity except by means of parenchymatous tissues. 

 Thus here at any rate the axes existed first before their steles. 

 Continuity between the steles of members of the different orders 

 of axes was a later modification, the usefulness of which soon 

 became obvious. 



We may picture Psilophyton as a terrestrial plant growing 

 under damp conditions in a bed of peat, and no doubt in such 

 a habitat a partial, little-developed and discontinuous con- 

 ducting system met the case. 



The next stage in the evolution of the stele appears to have 

 been the substitution of a continuous for a purely initial trans- 

 formation of proeambial elements. When the first set of elements 

 had been converted (protoxylem) the process continued and 

 primary wood, focused on the single protoxylem, was evolved. 

 This stage is thus the evolution of a monarch strand. Such a 

 stage exists for instance in the rootlet (not the rhizophore) of_ 

 Stigmaria. 



The next step appears to have been different in different cases. 

 The proeambial activity having ceased, a secondary cambium 

 was sometimes (but probably rarely) initiated and secondary 

 wood was added to the framework of a single monarch strand. 

 This state of affairs is also perceived in certain Stigmarian 

 rootlets. 



A more common and perhaps more successful course of stelar 

 evolution in stems was the substitution of a number of proto- 

 xylems with prolonged cambial activity in the central region 

 of the stem in place of a single such group. These groups may 

 have been centric or excentric, in the latter case the groups being 

 concentric. 



In the case of centric groups of protoxylem, space was naturally 

 limited, and even where the groups were excentric but con- 



