E. M. Kindle — Fossilization of Palceozoic Tree-trunks. 339 



Another section of a small birch-tree represented by perfectly pre- 

 served bark from Avhich the wood had wholly disappeared illustrates 

 well the possibilities which the birch affords for the formation of 

 sand-filled fossil tree-trunks similar in their general features to those 

 of the Palaeozoic Lycopods. This was found nearly filled with sand 

 and partiallj' buried in a river bar. The density of the bark of the 

 birch is such that a section of it remains almost completely submerged 

 when in the water. This characteristic would facilitate its speedy 

 sinking and burial in sediment whenever subjected to fluvial action. 

 The birch-bark specimen shown in the figure has suffered a slight 

 flattenino;, resulting from the decay and removal of the supporting 

 wood. It sliould be noted that this slight flattening of the birch-bark is 

 duplicated very often in the fossilized trunks of Zepidodendro?i. The 



Section of the bark of the trunk of the canoe-birch from which the wood 

 has completely decayed, but leaving the bark in a nearly perfect state of 

 preservation. 



partially sand-filled section of birch-bark just alluded to clearly 

 illustrates how perfectly one of the trees of our present flora combines in 

 itself characteristics which would lead to the same kind of fossilization 

 so often met with in Lepidode^idron. Obseryations of the empty bark 

 shells of decayed birch-trees in the Canadian forests can hardly fail to 

 convince one that Lepidodendron possessed a bark which, like that of 

 the birch, far outlasted the woody interior. It appears most probable 

 that the bark cylinders of the fossil Carboniferous Lycopods were 

 generally hollowed out by rapid decay of the wood and then filled 

 with sediment before fossilization began, just as empty birch-bark 



