XVIII] COMPARISON 37 



Putting aside certain anomalies, such as Stanropteris and Astei-opteris 

 which will be specially considered later, the increase in diameter of the 

 petiole runs substantially parallel with the increasing complexity of its 

 vascular tract. The measurements are here based chiefly upon the xylem, 

 and must be held as approximate. The series starts as before from Botryo- 

 pteris with its simple oval meristele of small size. With slightly larger 

 dimensions the Clepsydroid type appears having a dumb-bell-shaped section. 

 Where the dimensions are still larger this type is elaborated in various detail, 

 with formation of the arms or "antennae," and finally they are seen to arch 

 into an almost complete circle in the largest of all, viz. Ankyropteris with its 

 striking double-anchor outline (I^g. 326). Meanwhile the ratio of size of 

 petiole to size of the vascular tract gives no very distinct result till the last 

 examples are reached, in which the trace is seen to have expanded in much 

 higher ratio than the petiole that contains it. A word is due on the relatively 

 late Botiyopteris forensis of Permian origin: for its meristele is much larger 

 than in other species, and it presents the characteristic three points which are 

 specially deep in the large specimen figured by Bertrand. It is significant 

 that this is the latest, the largest, and the most complex of the known species 

 of Bot7yopteris. It would indicate that a phyletic increase in size bears with 

 it increased complexity of vascular structure, and this is consistent with a 

 like interpretation for the Zygopterids. 



The ancient Asteropteris appears to be anomalous anatomically. It has 

 one of the largest stellate xylem-masses (16 mm. in diameter), but the xylem 

 was undifferentiated. The leaf-traces were, however, relatively small 

 (2-3 mm.), and of the Clepsydroid type. This disproportion it shares with 

 Asterochlaena, which has also a Clepsydroid leaf-trace. It seems probable 

 that these two markedly stellate types show a structure consequent on 

 enlargement of the axis and stele without any proportional enlargement of 

 the leaves which it bears. In particular this seems probable for Asterochlaena, 

 since there is in it no constant relation of number or position of the leaves 

 to the flanges of the stele. Here stellation and phyllotaxis have not coincided. 

 This discrepancy is readily intelligible if stellation be an independent con- 

 sequence of increase in size. (See Proc. R. S. E. Vol. xliii, p. 117, 1923.) 

 Other early Fern-like plants still imperfectly known, of which Cladoxylojt is 

 quoted in the table as a single example, suggest that in them also decen- 

 tralisation and even disintegration of thevascular tract followed upon increase 

 in size. (Solms-Laubach, AbJi. d. K . Preuss. Geol. Laudesanstait, Heft 23, 

 1896; P. Bertrand, Progressns, IV, ii, pp. 249-254,) 



Lastly, Tidncaiilis and Aiiachoj-optcris, with their relatively large and simple 

 meristeles in the petiole (3'2 mm. and 4-6 mm. diam.), appear to be also 

 anomalous, other traces of similar dimensions being Cleps}-droid. Their 

 margins are revolute so that the convexity is directed adaxially (Fig. 315). 



