378 STUDIES IN FOSSIL BOTANY 



Lyginodendron oldhamium, is clearly a point of great 

 importance, for from the characters of the stem alone 

 the affinity with Ferns is much less evident. This 

 proof has been afforded, first, by finding the petioles in 

 continuity, through their finer ramifications, with the 

 Sphenopteris leaflets, and next, by finding the petioles 

 themselves in continuity with the stem. The petiole is 

 connected with the stem by a broad but not sheathing 

 base (see Fig. 138). The twin leaf- trace bundles bend 

 out sharply into the leaf-base, sometimes becoming 

 fused as they enter it, sometimes remaining distinct. 

 The sclerotic nests, already referred to, are very abundant 

 in the cortex of the axillary region. The petioles 

 found in connection with the stem can often be traced 

 for some distance after they become free, and their 

 branching can be followed ; sometimes a fortunate 

 section passes through the point of bifurcation of the 

 main rachis. There is thus no doubt that the leaves of 

 Lyginodendron oldhamium were of a forked and highly 

 compound form. In all their characters, so far as the 

 vegetative condition is concerned, they suggest the foliage 

 of a Fern (see below, p. 384). 



3. The Roots. — The adventitious roots of Lyginoden- 

 dron oldhamium were known for many years as separate 

 objects, before their relation to Lyginodendron was 

 discovered. Like so many other detached organs of 

 fossil plants, they received a distinct name, and were 

 long known as Kaloxylon Hookeri. They have since 

 been so constantly found in connection with the stem, 

 that their true nature is fully established. 



The roots are often most perfectly preserved, and 



