THE FOSSTL OSMUND AC EtE. 



459 



distance into the xylem of the stem so as to form a little pocket. This, however, soon 

 dies out below, leaving the protoxylem of the trace deeply immersed in the solid xylem 

 strand of the stem (fig. 12). Lower down the protoxylem itself dies out. Fig. 14, 

 PI. II., is a photograph of a longitudinal section of an actual leaf-gap. It is not, how- 

 ever, quite vertical, and in consequence the internal wedge of xylem is probably some- 

 what exaggerated in extent. 



All this may be expressed in another way by saying that in Osmundites Kolbei the 

 departure of a leaf-trace does not at once cause a break in the xylem ring, whereby it 



Fig. 3. — Osmundites Kolbei. Diagram of a vertical section through the departure of a leaf- trace from the xylem ring. The 

 xylem ring is not at once broken through. In this case only a very shallow parenchymatous pocket is formed. Contrast 

 with text fie. 2. 



differs from the ordinary procedure in the modern Osmundacese. At the same time, 

 leaf-traces are sometimes to be met with in Todea hymenophylloides, which depart 

 exactly as described in Osmundites Kolbei. 



Although the leaf-trace departure described above is by far the most frequent in 

 Osmundites Kolbei, still there is considerable variation in detail. For instance, the 

 distance to which the plate of xylem extends above the actual departure of the leaf- 

 trace xylem may be longer or shorter. Again, the decurrent pocket of soft tissues may 

 be very short or entirely absent (text fig. 3 ; compare also figs. 8 and 9), in which case 

 the protoxylem of the leaf will almost at once become completely immersed in the 

 xylem strand of the stem. It is possible even that departures of the ordinary Osmunda 

 type may occur, although such were not observed in our sections. 



It appears, therefore, that, as regards the departure of the leaf-traces, Osmundites 

 Kolbei is transitional between Osmundites Dunlopi, where they leave no break at all 



