2i 4 STUDIES IN FOSSIL BOTANY 



be taken as representing the Syringodendron surface, 

 though on a small scale. In typical Syringodendron 

 the parichnos- prints may be enormously enlarged, 

 reaching a length of more than a centimetre. 



In some cases the leaf-scars on the Sigillarian surface 

 are accompaned by other scars, usually regarded as 

 marking the insertion of the fructifications (see Fig. 87, 

 c.s.). We will, however, postpone the consideration 

 of these, until we come to deal with the cones them- 

 selves. 



As regards the habit of the Sigillariae, our knowledge 

 is still very imperfect. In some of the forms, forking 

 branches have been observed, though they are not as a 

 rule common ; in stems with the Rhytidolepis type of 

 sculpturing no branching has yet been observed. In 

 the unbranched forms the habit appears to have been 

 something like that of the Australian Grass -tree, 

 Xanthorrkoea, the tall, upright shaft terminating in a 

 sheaf of long, grass-like leaves. In other cases, as in 

 the Favularian species 5. elegans, and in S. Brardi, of 

 the ClatJiraria group, the stem divided, by successive 

 dichotomy, into a few large branches. 



2. Anatomical Structure. — Recognisable specimens 

 of Sigillaria with structure preserved are comparatively 

 rare ; of late years, however, several new examples have 

 come to light. It is possible that the rarity of 

 specimens with structure, while the structureless casts 

 are so common, may be in part due to the smaller 

 branches not having always been recognised as belonging 

 to Sigillaria. The fact, however, would also find an 

 explanation, if branching occurred very sparingly, for, 



