<i72 



fusing witli the peliolar bundle. At a somewlial lower level the 

 continuity of the peripheral loop formed bj this fusion is disturbed. 

 The interruption thus formed is limited on both sides by the down- 

 ward continuation of the halfs of this peripheral loop. The xylem 

 of the next pinna-bar fuses with the two ridges at its extremities. 



In E. Scotli we see the pinna-bar approaching the petiolar bundle 

 too. But just before its fusion with the latter it is split up in its 

 middle part; thus two separate bundles are formed, which fuse 

 with the petiolar bundle. We see here the same fusion with the 

 petiolar bundle and the same interruption in the pinna-bar; but in 

 E. Bertrandi the highest point of the interruption is below the 

 fusion of the pinna-bar with the petiolar l)undle and in E. Scotli 

 it lies above this point. 



The interruption, the height of which is different in these two 

 species, is always limited below by the next pinna-bar. It lies above 

 the insertion of the pinna-bar. The relative length of the interruption 

 to the distance between two pairs of pinnae determines the condition 

 of the transverse section. In E. Scotti the distance between two 

 successive pairs of pinnae is l)ut small, often the bundles of two 

 pairs of pinnae are shown on tlie same side in one and the same 

 transverse section. 



Tlius the structure of Etapteris Bertrandi Scott enables us to 

 explain the features in other more complicated species of Etapteris. 

 On the other hand it has manj' points in common with simpler 

 forms, e.g. Diplolabis Römeri (Solms) Bertrand. In this plant an 

 interruption above the insertion of the pinna-bar is present too. 



If tiie petiolar bundle is followed here in its downward course, 

 which Gordon's') researches enable ns to do, it can be shown, that 

 the lowest pinna-bar encloses at its inner side an amount of paren- 

 chyma by the fusion of the pinna-l)ar with the two sides of the 

 interruption. At a lower level the two protoxylems which are situated 

 on both sides of the parenchyma fuse. The parenchymatous tissue 

 diminishes in size and ends blind below. 



But throughout its course to its lowest point it is in contact with 

 the protoxylem; it seems as if the lowest part of the parenchymatous 

 tissue follows the course of the protoxylem when penetrating into 

 the tracheides of the metaxylem. 



It is remarkable that in these plants the protoxylems are always 

 associated with parenchyma except in the lowest part ; this paren- 

 chyma, or at any rate thin-walled tissue, is situated at the adaxial 



1) W. T. Gordon, 1911. 



