THE STRUCTURE AND AFFINITIES OF DIPLOLABIS ROMERI (SOLMS). 713 



on the shore, or, as has been demonstrated,* embedded in the overhanging cliffs. 

 Some of my specimens were obtained from loose blocks, while others occurred in situ in 

 the agglomerate which forms the cliff near Pettycur House. I have collected two large 

 blocks (one being about 1 ft. 6 ins. x 1 ft. x 1 ft., and the other 1 ft. x 9 ins. x 6 ins.), and 

 a small rounded fragment about 6 ins. diameter, containing petrified specimens of this 

 plant. I have further been able to obtain numerous detached petioles and pinnae of the 

 fern from the agglomerate at Pettycur House. All these specimens, with the exception 

 of one of the first-mentioned blocks, were obtained in situ in the cliff. The plant 

 petrifactions were more or less continuous in the blocks, and thus it was possible to 

 prepare long series of sections illustrating the structure of the various members of the 

 plant. These sections were cut approximately at equal distances from each other, the 



N? 



Stage oF 

 PI. ii Fig. 18 



Stage oF 

 PI. ii Fig. Zl 



Stage oF 

 PI. ii Fig. 24 



STage oF 

 /','. //' Fig. 25 



STage oF 

 PI. ii Fig. 26 



Stage oF 

 Ft. II rig. 28 



Slides 



i 















701 - 711 







2 















713 - 733 









3 















650 -700 









4 















284(a-z)-294 











5 















340-356 

 & 284- 284 p 











6 















335-338 







7 















331-334 







Fig. 1. — Diagram illustrating the amount of the overlapping in the various series of sections cut from the 



petiole-trace of Diplolabis romeri. 



interval between the sections being from ^ in. to ^ in., though in some cases it did 

 not exceed -^ in. In general, the changes in the specimen from section to section are 

 not rapid. The slides examined, which number about 250, were prepared either by 

 myself or by Mr F. Phillips of the Sedgwick Museum, Cambridge, under my super- 

 vision. From one petiole 55 slides were prepared in series, but even that number was 

 not sufficient to illustrate completely all the stages in the change of form of the petiole- 

 trace. The various series which were prepared overlap so much, however, that there is 

 no doubt of the stages described below, each series serving to check the others. The 

 extent of the overlapping is indicated in text-fig. 1. 



The mode of preservation of most of the specimens is somewhat peculiar. The 

 xylem elements are generally clearly preserved, but the thin-walled tissues of the cortex 

 are not as a rule well petrified. Even in the xylem elements the crystallisation of the 

 calcite has often distorted the walls. In the softer cortical tissues, the spherules of calcite 



* Gordon, Trans. Geol. Soc, Edinburgh, vol. ix. pt. iv., 1909. 



