732 



MR W. T. GORDON ON 



indicated that a strong family likeness exists between the Osmund aceas, Zygopteridcze, 

 and Botryopter ideas. The discovery of the stem of Diplolabis has still further 

 strengthened this general position, as it indicates that evolution in the zygopterid stem- 

 xylem was along the same lines as that shown in the Osmundaceas. 



In their petiolar structure the BotryopterideaB seem to stand much nearer the 

 Osmundacese than are the Zygopteridew. In Botryopteris antiqua Kidston, for 

 example, I find that the petiole-trace has a single sunk protoxylem group which lies 

 adaxially in the trace and soon comes to lie adaxially on the periphery of the trace. In 

 its subsequent divisions the trace gives off primary pinnae in two orthostichies. This is 

 quite similar to the osmundaceous trace, which in general has a sunk protoxylem group 

 on the adaxial side. Gradually the trace opens up on the adaxial side as we ascend 



Pro to clepsydropsis 



Si 



S 



! 



3 



Clepsydropsis 



Dineuron 



MeTa clepsydrop sis 



1 



Efapteris 



Diplolahis 



Zygopferis 



Fig. 4. — Table illustrating the relation of Diplolabis to the other Zygopterid ese in which two pinna-traces are 



derived from a single pinna-trace-bar. 



until it becomes horseshoe-shaped ; the protoxylem strand has meanwhile divided into 

 several branches, all lying on the inside of the horseshoe-shaped trace. Here, again, 

 there are two orthostichies of primary pinnae. 



The Zygopterideas, on the other hand, have two sunk protoxylem groups in the 

 petiole-trace, and two or four orthostichies of primary pinnae. In the cases where there 

 are four orthostichies, the arrangement of the pinnae on the rachis would not appear to 

 have been very successful, for the pair on one side must necessarily shade the pair on 

 the other side, if the petiole were held flat ; if the petiole were held erect, as was the 

 case in D. ro'meri, the higher pinnae would shade the lower. 



The discovery of Kidston and Gwynne-Vaughan, that several traces in their speci- 

 men of Thamnopteris schlechtendalii possessed two protoxylem groups, is exceedingly 

 interesting as connecting the osmundaceous leaf-trace with the zygopterid type of trace. 

 Thus Thamnopteris schlechtendalii seems to indicate an ancestral relationship with the 

 Zygopteridew. 



In stem structure the Botryopteridese are certainly more primitive than either oi 



