THE FOSSIL OSMUNDACEyE. 653 



however, becoming conspicuously small, and even the outermost possess the same kind 

 of pitting as the rest. Elements with small lumens also appear to be scattered 

 throughout the outer xylem, but longitudinal sections show that these are really due 

 to the transversely cut ends of the tapering tracheae. The only definite protoxylem 

 elements to be observed in the stem stele are those decurrent from the leaf-traces which 

 form more or less deeply immersed groups at various points in the ring of outer xylem 

 (fig. 6). These protoxylem strands consist of small tracheae with a single series of 

 scalariform pits on their walls, and they die out gradually below without joining on 

 to one another. 



The central xylem is sharply marked off from the outer xylem ring both as seen 

 in transverse and in longitudinal section (figs. 6 and 7). The strong contrast it 

 makes with the outer xylem is due to the fact that it consists of short and wide 

 elements very irregular in shape, but on the whole more or less vesicular or sack-like 

 (fig. 11). In transverse section some of these elements appear to be thinner walled 

 than the rest, in particular towards the central region ; but longitudinal sections show 

 clearly that the whole of the central xylem consisted originally of entirely similar 

 tracheal elements, and that these apparent differences are the result of variation in the 

 amount of decay undergone by the parts in question before mineralisation. The short 

 tracheal elements of the central xylem were no doubt thinner walled than those of the 

 outer xylem ring, and possibly also less lignified, and they further differed in the 

 nature of their pitting. The distribution of their pits is completely irregular, and they 

 can only be described as reticulate (figs. 12 and 13). It is instructive to contrast these 

 figures with those of the more regular "porose" tracheae of the central xylem of 

 Zalesskya gracilis shown in figs. 5, g and h, PL II., in Part II. of this series. All 

 the elements of the xylem, both central and peripheral, are really vessels, the pits 

 being actual perforations. In the reticulate elements each pit represents a distinct and 

 separate perforation ; but when the pits are in regular vertical series all the pits of one 

 series communicate with one another by a split or empty space in the substance of 

 the wall, as is also the case in the other Osmundaceee both recent and fossil.* In 

 Thamnopteris the xylem is preserved in such a way that the holes in the middle of 

 the wall appear as light-coloured spaces enclosed in dark-coloured bars that represent 

 the solid regions, giving an effect similar to that we have already figured in Osmundites 

 skidegatensis (Part II., PI. II. fig. 7). 



We have no doubt that in the living plant the xylem was quite solid and con- 

 tinuous up to the very centre, although in our sections there is a small central area 

 unoccupied by tissue. Indeed, in the section represented by fig. 6 it is evident that 

 if the more central elements were freed from contraction and restored to their 

 original form and dimensions they would suffice to completely fill up the area 

 in question. 



* D. T. Gwtnne-Vaughan, " On the Real Nature of the Tracheae in the Ferns," Ann. Bob, vol. xxiii. (No. 87), 

 pp. 517-523, PI. xxviii., 1908. 



