4 Weiss, A Tylodendron-Uke Fossil. 



tinuous passages passing vertically between the cells of 

 the pith, and in this view, too, they resemble the mucilage 

 canal of the Cycadales. It will be seen, however, that in 

 our specimen of Tylodendron the duct has usually no 

 specialised secretory cells of smaller size than the ordinary 

 pith cells. It is, therefore, of a less highly organised or 

 specialised nature. In some cases, as illustrated by PL II , 

 Fig. 7, some of the surrounding cells may show that a 

 division has taken place, though this is not the usual con- 

 dition. In one case {PI. II., F'ig. 4) repeated division has 

 taken to form a very specialised secretory and surround- 

 ing layer. 



These secretory passages form a very striking feature 

 of the pith, and are certainly ver}' distinctive of the Tylo- 

 dendron under consideration. They are not found 

 throughout the pith but are numerous in the outer layers 

 to the depth of about i cm. from the periphery. Their 

 course, as seen in longitudinal sections, is vertical, without 

 bending about as is sometimes the case with similar ducts. 

 I have not seen them passing out through the foliar gaps, 

 nor do they seem to have an}- connection with the leaf- 

 traces. Running vertically as they do, they do not show 

 clearly any anastomosing, but sometimes two may be seen 

 running closely side by side ; and from such transverse 

 appearances as are shown in PI. //., Figs. 2 & 3, one may 

 conclude that at times the}' unite after running side by 

 side for some distance. In no case have secretory passages 

 of any kind been found among the wood}' tissues of the 

 plant. 



The pith cells are all parenchymatous in character, 

 and, as is seen most clearly in longitudinal sections^ slightly 

 drawn out in a horizontal direction, except at the outside, 

 where they join on to the vascular tissue. Here their 

 long axis is parallel to that of the adjoining tracheids. 



