129 



that desquamation is any proof of its existence in 

 this instance more than in that one. 



Either in the case of simple rounding, or of 

 desquamation, the cuboidal, or prismatic form, 

 becomes ultimately spheroidal; for mechanical 

 reasons sufficiently obvious. But as instances 

 also occur where a simple prism exhibits more 

 than one desquamating centre, it is undoubted 

 that, in some cases at least, there is a lamellar con- 

 cretionary structure united to the prismatic. 



In the columnar structure, as the term im- 

 plies, the prisms generally possess a considerable 

 length in proportion to their breadth, and they 

 are not limited to the quadrilateral form. In a 

 few instances, from the extreme shortness of the 

 prisms, the columnar passes to a tabular, or a 

 lamellar and jointed structure ; but the two are 

 united under this head, in consequence of their 

 general resemblance, and of the undefinable line 

 by which they are separated. Where the 

 columnar structure is on a very large scale, it 

 might be supposed to belong to that division 

 which is included under the term of external 



K 



