HITHERTO KNOWN AS STERNBERGIJE. 35 1 



outline. Many of the lamin89 subdivide on one side into 

 two — a small incomplete space, not extending across the 

 entire medullary canal, being thus intercalated. 



Now, all this is precisely what has occurred in the so- 

 called Stembergiae, only the process of absorption has been 

 carried a step further than in any of the recent examples 

 already quoted. In addition to the formation of the thin 

 transverse disks, the centre of each disk has usuall}- been ab- 

 sorbed, leaving merely flattened rings of medullary tissue, 

 united to the rest of the pith by their thickened peripheries. 

 Thus, instead of the intermediate cavities being isolated, 

 they have apparently been connected together by a central 

 fistular passage. The appearance presented by a vertical 

 section of the pith, when recent, would apparently corre- 

 spond with fig. 11, which is a restored sketch of what I 

 suppose this structure to have been. The gradual thicken- 

 ing of the exteriors of these rings would give a concave outline 

 to each cavity when intersected vertically, which would of 

 course be represented in the cast by the rounded outline 

 seen in the peripheries of what in Stembergia were thought 

 to be horizontal plates ; and the divergence of many of the 

 laminae accounts for the apparent intercalation of incomplete 

 disks, in the fossil, between those which can be traced round 

 its entire circumference. Examples of this feature are seen 

 in fig. 10. Well might botanists wonder at the anomaly of 

 vegetable structures assuming this strange arrangement of 

 "*' horizontal plates, held together by some connection in the 

 axis of the stem,'' and add, " a most extraordinary appear- 

 ance, to which we know of no parallel." — {Lindley and 

 Ilutton's Fossil Flora of Great Britain, vol. iii., p. 188). 

 The anomaly is now explained ; this wonderful arrangement 

 of vegetable tissues has no real existence. 



In the Coalbrookdale specimen (fig. 1), the transverse 

 ridges are exceedingly numerous, and in very close con- 

 tact. In this respect they differ a little from the ordinary 



