254 Dr. Mac Culloch on a 



exposing various trap rocks, as, for example, basalts, of suffi- 

 cient bulk, to a heat short of that required to fuse them, and 

 for a sufficient length of time, to be determined only by repeated 

 examinations. Should such a result take place as that which 

 has occurred in the sandstones, we should then be in possession 

 of a valuable fact from which to deduce a train of geological 

 reasoning respecting the figured traps, which must now be 

 supplied by conjectures and analogies. What may be offered 

 on that subject I shall now defer, till the natural appearances, 

 which resemble the artificial one already described, have been 

 detailed. 



I shall now, therefore, proceed to describe some appearances 

 in nature, which appear not only to illustrate the preceding 

 experiment, if such it may be called, but to show the con- 

 nexion which it has with similar phenomena in the structure of 

 rocks. That connexion it must, however, be admitted, hangs 

 on a hypothetical view of a cause ; or rather, on an inference 

 regarding the probable action, in this particular instance, of a 

 certain state of things which is now almost universally admitted 

 to have existed. But the nature of the reasoning on this 

 subject will be better understood when the appearances them- 

 selves have been described. 



That substance known by the name of " columnar ironstone" 

 is probably familiar to most mineralogists. It is found under 

 many different modifications, and these principally relate to 

 the sizes of the prisms of which it is composed. For the 

 objects of the question in view, although it is not useless to 

 inquire respecting the causes of this peculiar configuration, it 

 is impossible to reason very satisfactorily from it, unless the 

 geological situation of all such specimens were carefully 

 examined. Unfortunately, the only instances with which I am 

 practically acquainted are those which occur in Arran, which 

 I have described in my work on the Western Islands of Scot- 

 land. 



These examples are noted for the large size of the prisms, 

 and for the facility with which they separate into short joints ; 

 while, in one of them, there is further a peculiar character 

 produced by a groove or channel on the ends of the prisms, 

 and parallel to their sides. These peculiarities, however, seem 



