CYATHEAE 609 



with the main stele, except sometimes by a small branch near their point 



of origin. 



It seems, therefore, that the internal vascular strands of Alsophila 

 xcelsa owe their existence to the same initial phenomena as do those of 



Dennstaedtia rubiginosa ; that is to say, they are probably derived from 

 lie elaboration of a local thickening of the xylem-ring at the margins of 

 leaf-gaps in the ordinary stelar cylinder ; but they do not appear at 

 11 until the ordinary stelar cylinder has become dictyostelic. 



The ontogeny thus disclosed for a complex Tree-Fern may be held as 

 valid suggestion of the way in which the mature condition was achieved 

 n descent. It starts from a protostelic state, which is, however, brief, and 

 )asses to the solenostelic by intrusion of outer-lying tissues into the xylem- 

 ore ; but this again passes into the dictyostelic by reason of the overlapping 

 f the leaf-gaps : and lastly, by intrusion of vascular growths from the 



margin of the leaf-gaps, the medullary system is produced. All these 

 teps, so quickly passed over in the individual life, are readily intelligible, 

 ,nd even probable, in the evolutionary story of plants with a massive 

 xis, bearing large and closely disposed leaves. 



The protostelic state, here so short, is the permanent condition in 



most of the Gleicheniaceae. But the most advanced species of Gleichenia 

 G. dichotoma and pectinata) show signs of solenostely, while in Alsophila 

 mrinata the solenostelic state appears to be permanent. But in other 

 pecies of Alsophila it also is a phase quickly passed through to the 

 iictyostelic state, which is then permanent. Finally, the medullary 

 ystem absent in A. pruinata, as it is also in Dicksonia, but developed 

 n Cyathea, is clearly a late accessory, probably consequent upon the 



enormous distension of the pith in relation to the wide leaf-bases. 



The leaf-trace also presents features of comparative interest : in the 

 plant it consists of a single strand, as it is in all the Gleichenias : 

 n A. pruinata it appears to be so at the base even of the mature leaf, 



though it soon breaks up into separate strands as it . passes up into the 

 eaf-stalk : but in most of the Cyatheae the leaf-trace in the mature 

 shoot is from the first composed of a number of distinct strands. These 

 uccessive steps again indicate a probable phyletic progression, the young 



plant showing a condition similar to that seen in simpler types, and 



especially in the Gleicheniaceae. 



If the facts derived from the characters of the sorus be put into 

 relation to these from anatomy, a substantial parallelism' emerges, point- 

 ng in both cases towards the Gleicheniaceae as a probable indication 

 of the genetic source. In soral characters Alsophila is the nearest to 

 Gleichenia, and especially to those species in which the sorus is no 

 Conger uniseriate, but consists of a large number of relatively small 

 sporangia (G. dichotoma and pectinata). It is in these very species that 

 there is a definite advance towards a state of solenostely not very far 

 removed from that actually seen in A. pruinata. From such a 



2 Q 



