296 THE BORDERLAND OF SCIENCE. 



they remained erect after the sea had made its way into 

 the domain of the ancient forest? Many geologists 

 think so ; and doubtless the stumps of stout trees 

 might resist for a long time the action of the sea 

 waves. But there seems good reason for believing that, 

 when the submergence first took place, these stumps 

 stood but little above the upper surface of the vegetable 

 layer, or that in many instances the trees were broken 

 off even below that level. Then, as the pressure of the 

 superincumbent layer gradually increased with the 

 layer's increase of thickness, the vegetable matter was 

 pressed down below its former level, and the stumps 

 were left standing above the depressed surface of the 

 vegetable layer. This explains the conversion of the 

 bark of these stumps into coal, since there is every 

 reason to believe that stumps simply left imbedded in 

 sandstone would not change into pure coal. 



In passing, I may remark that in whatever way it 

 happened that the stumps of the ancient forest-trees 

 remained standing above the level of the vegetable 

 mass forming the coal-seam, a strange result has 

 followed. The upper part of the stem became filled, 

 as I have said, with sandstone, forming a cast of the 

 interior of the ancient tree ; the bark became coal ; and 

 outside the bark is sandstone again. Thus there is a 

 mass of sandstone separated from the surrounding 

 sandstone by a tube of coal. This mass is not cylin- 

 drical, being larger below than above ; so that if in any 

 way the mass ceases to be supported, it falls like a 

 bolt from a gun. But. in working the coal-seam the 



