758 PROFESSOR W. C0. WILLIAMSON AND DR. D. H. SCOTT ON THE 
he more central cells of the cylinder are relatively large and may well represent 
undeveloped tracheze. Immediately outside the latter we find groups of small cells, 
possibly young phloém. The outer layers of the wide cortex are large-celled, as in 
roots of the “ Kaloxylon” type. Numerous “secretory sacs ” are present throughout 
the tissues of the organ. We see no other explanation for this structure than that 
it represents a young adventitious root, and, if so, it is of some interest, for organs 
at such an early stage of development are necessarily rare in the fossil state. The 
presence of such a rudimentary root in a mature stem is not very surprising, for we 
know that adventitious roots may arise at all ages of the stem, and also that such 
roots may remain for a long time undeveloped.* 
2. Structure of the Root. 
Our knowledge of the structure of the root in H. Griev is at present a matter of 
inference rather than of absolute proof, for we have no clear transverse sections of 
those roots which are in actual connection with the stem. The longitudinal sections 
of the root-bases show that the xylem extended to the centre of the cylinder, and 
mainly consisted of tracheides like those of the stem. 
In the Burntisland preparations transverse sections of roots are rare and unsatis- 
factory. The Dulesgate specimens, however, contain many beautifully preserved 
roots, which in some of the preparations are seen in trausverse section in close contact 
with the same stems on which the bases of adventitious roots are found. All these 
roots are of the “ Kalozylon” type, and, as the same material contains stems of 
Lyginodendron, it is quite probable that some of them belong to the latter plant. 
Yet we feel sure that among these roots those of Heterangium are included, for we 
know that the stems of Heterangiwm in this material produced roots freely, and there 
are no other root-like organs present which could possibly be referred to them. We 
believe the explanation to be that the roots of Heterangiwm G'rvevit were themselves 
of the ‘‘ Kaloxylon” type, and not always readily distinguishable from those of 
Lyginodendron. 
There is one special form of root, however, which we feel pretty strongly convinced 
belonged to Heterangium Grievit. We refer to certain tetrarch roots, such as that 
shown in transverse section in Plate 27, fig. 27. Roots of exactly this kind are 
excessively common in the preparations from Dulesgate material containing Hete- 
rangium Grievit, while we have never found them associated with Lyginodendron 
when Heterangium is absent. The characteristics of these roots (which, as the figure 
shows, are often perfectly preserved) consist in the large tetrarch primary wood, with 
very prominent angles, and in the almost geometrical regularity with which the tissues 
are arranged. Roots of Lyginodendron with a tetrarch cylinder are, it is true, well 
* The best preparations illustrating the connection between root and stem in H. @rievii are O.N. 1260, 
1294, 1915, 1915 A, 1915 G, and 1915 H. 
