FOSSIL PLANTS. 485 



the irregular fragments of silicified wood, found in connection with our recent 

 formations, and which, in some countries in Arkansas and Mississippi, for ex- 

 ample are in some places strewn upon the ground in profusion. Agglom- 

 erations of silex are rarely homogeneous or regularly compact throughout. 

 They are interspersed with fissures or soft veins which, when penetrated by water, 

 expand under the influence of frost, and determine fractures in various direc- 

 tions. But fossil wood broken in that way is rarely found in our Carboniferous 

 measures. Generally, the fossil trees of this formation, when separated from 

 the mineral substances in which they were originally imbedded and petrified, 

 show the fracture by horizontal divisions, as by cleavage, and when in a stand- 

 ing position, and taken out of the matter which surrounds them, they separate 

 in disks of various lengths, and can thus be taken out in pieces, which super- 

 posed afterwards rebuild the whole trunk, without marks of any other mode of 

 disconnection, but horizontal through fissures. In that way the different parts 

 of the trees mentioned above, as found by Dr. D. D. Owen, have been taken 

 out of the sandstone separately and replaced in their order of superposition, to 

 rebuild the vegetable in its original position. AtCarbondale, in Pennsylvania, 

 a true forest of Catamites has been crossed in the opening of an inclined tun- 

 nel through a bank of sandstone to a bed of coal underlying it. The fragments 

 of petrified stems taken out of this passage are in such abundance that they 

 have been used for the construction of a kind of gangway for running the coal 

 cars out of the mines. These fragments, nearly without exception, are mere 

 disks, varying in length from one to four inches, without relation to the size 

 or diameter of the stems, which measure from three to six inches; the differences 

 in the length of the sections being as marked for the large as for the small 

 stems. All these fragments represent only as far, at least, as I could determine 

 from the examination of hundreds of specimens, two species of Calamites : 

 G. SucJcowii and G. approximatus, Brgt. The walls of the tunnel are adorned 

 by a number of these trees, still in their standing position and half imbedded 

 in the sandstone. Though these stems are continuous, they show, at various 

 and irregular distances, horizontal fractures where they break or are dislocated 

 at their separation from the surrounding sandstone. Some of these trunks of 

 Calamites, which in their natural state were evidently hollow, have been 

 abruptly folded or crushed, like hollow cylinders in bending under their own 

 weight, or by some external force; but even at the point of inclination or tor- 

 tion of these stems, the fracture is horizontal or perpendicular to their erect 

 position. At Paintsville, Johnson county, Kentucky, the bottom of the river, 

 which at some places has been cleanly washed, is marked, as in a kind of irregu- 

 lar mosaic work, by the broken tops of large trunks of Sigillaria, still in their 

 original standing position, all horizontally fractured. One of these trunks 

 measures twenty-two inches in diameter. The same peculiar kind of horizontal 



