INTRODUCTION TO CRYPTOGAMIC BOTANY. 515 



the bark, and arranged in a circle symmetrically round the axis 

 with short interspaces, through which the other smaller bundles 

 dispersed in the central mass give off branches into the stipites, 

 while others exist in the space between the bark and larger 

 masses. These latter form a more or less perfect cylinder 

 round the axis, and are altogether distinct from anything in 

 Endogenous stems ; besides which, there is not that crossing 

 of the fascicles characteristic of Endogens. The vessels 

 themselves, which are always scalariform, though varying 

 greatly in size, are accompanied by cellular tissue, and sur- 

 rounded by elongated pale wood cells (pleurenchym), beyond 

 which is parenchym mixed with resinous cysts. The hard 

 coat which incloses the vessels with the pleurenchym and 

 parenchym, belongs to the general mass of cellular tissue, and 

 not to the wood which is represented by the minute quantity 

 of pleurenchym. The structure of other ferns is more or less 

 in accordance with this; but the disposition of the bundles and 

 of the hard attendant tissue is infinitely varied. In Pteris 

 aquilina, for instance, the hard tissue which in the tree 

 ferns incloses the principal vascular bundles, is disposed in 

 two curves, on either side of which the vascular bundles are 

 arranged, each inclosed by a thia coating of denser tissue than 

 the general mass. Sometimes these two arcs meet at one 

 extremity, sometimes at both, especially where a stipe is given 

 off. Besides the two main masses of dense tissue, there are 

 many scattered fibres. In other cases this peculiar dense tissue 

 is altogether scattered about in little fibres, like those of Pteris 

 aquilina, just mentioned ; in some, as in Olearia hirta, it is 

 either converted into or replaced by curious cysts (Fig. 1 10, e,/), 

 which display a fibrous structure. In many cases, as in Phy- 

 'matodes leiorhiza, where the bundles are arranged in one 

 principal circle, with a few outliers, it appears to be entirely 

 wanting. In this and many other ferns there appears to be 

 scarcely any distinct cortical layer, the transition to the 

 general mass of cellular tissue being almost imperceptible. 



677. Not only is there great difficulty in arranging ferns 

 satisfactorily, but it is even more difficult to determine the 

 limits of species. If large series are examined, it is asto- 

 33 * 



