xviii] ZYGOPTERIDEAE 27 



Sahni) ; secondly, those with four rows, two being on each side of it (Dineuroideae, Sahni). 

 Also a further distinction exists between those in which at the departure of a pinna-trace 

 the xylem-strand of the pinna forms a closed ring (Fig. 325, 7-9), and those in which it has 

 the form of a slightly curved band, which divides into halves (Fig. 325, 10, n) ; or the 

 division may actually precede the separation of the trace (Fig. 325, 12, 13). The Clepsy- 

 droideae include Bertrand's genera Aiikyropteris (Fig. 326) and Clepsydropsis (Fig. 325, 

 7, 8, 9). In all of them the pinna-trace departs as a closed ring, and the island extends 

 along the whole length of the arms, except in Clepsydropsis where the arms appear to be 

 undeveloped, as they are also in the "assumed primitive type" (6). The Dineuroideae may 

 be divided into two series according to the relative extension of the parenchyma-islands. 

 In the first of these the island extends the whole length of the arms, and the pinna-trace 

 is given off as a long continuous band : as in Bertrand's genera Metadepsydropsis and 

 Diplolabis, and in Z. priinaria (10, n). In the second series the island hardly extends 



Fig. 327. Etapti'vis Scotti. Transverse section of rachis. In 

 the main bundle, the sieve-tubes on either side of the middle 

 band are preserved. On the left the two traces are entering 

 the common base of the paired pinnae. On the right a pinna- 

 bar is seen, (x 9.) S. Coll. 2009. (From a photograph by 

 ]Mr W. Tams. After Scott.) 



beyond the lateral termination of the cross-bar, and the pinna-trace is usually already 

 double at or near to the point of origin. This includes Bertrand's genus Etapteris (12, 13, also 

 Fig. 327). Lastly Dineuroti is related to this group (14), but it shows a simplicity Uke that 

 of Clepsydropsis. These genera last named may be held as representing a primitive type 

 from which the more elaborate types were derived ; and it is significant that that type 

 exists in the ancient Clepsydropsis aniiqiia (Fig. 328), and with slight modification in the 

 Devonian Asieropteris (Fig. 324). It thus appears that the Zygopterid leaves advanced far 

 along special lines of development of their own, and that they culminated in the production 

 of two rows of appendages on each side of the rachis. This last condition may well be 

 accepted as the result of a very early dichotomy of each pinna of the single-rowed type. 



Notwithstanding the good state of preservation of the rachis, and often of the axis of 

 the Zygopterideae, the knowledge of the upper regions of their leaves is remarkably im- 

 perfect. 



