224 STUDIES IN FOSSIL BOTANY 



itself, which was also thoroughly investigated by M. 

 Renault. 1 



The leaves with their structure preserved are in 

 association, but not in actual connection with the 

 stems. In cases, however, such as that illustrated in 

 Fig. 92, B, where the base of the leaf itself shows a 

 characteristic scar, corresponding with those on the 

 stem-surface, there can be no doubt as to the attribu- 

 tion. The general form and structure of the Sigillarian 

 leaf (the species chiefly studied was 5. Brardi, with its 

 variety, 5. spinulosa) is much like that of a Lepido- 

 dendron. Here also there are two deep longitudinal 

 furrows on the lower surface, one on either side of the 

 midrib, and it is only on the surface of these furrows 

 that the stomata are found (see Fig. 92, C, g). The 

 single vascular bundle appears to have been concentric 

 in structure, the xylem having its spiral elements 

 placed laterally. Below the bundle itself is a double 

 layer of sclerenchyma, beyond which is a band of 

 delicate parenchyma with scattered tracheides. This 

 accessory xylem may be regarded as corresponding 

 functionally to the transfusion-tissue of Coniferous leaves. 

 We have already described the same tissue in the 

 leaves of Lepidodendron, where, however, the extra- 

 fascicular tracheides are more abundant, and form a 

 complete zone round the bundle. The transfusion- 

 tissue of the Sigillarian leaves was at first regarded 

 by M. Renault as secondary xylem, but he after- 

 wards withdrew that view, considering that the true 

 secondary wood is limited to the leaf-trace, where it 

 passes through the cortex of the stem. It seems, 



1 Flore fossile d'Autun et d'lZpinac, Part ii. p. 213, Plate xli. 



