Walter Keeping — On Columnar Sandstone. 439 



directions of the columns in the several tiers of prisms. The 

 wondrous regularity of the columns and their small size are 

 particularly striking in this exposure, viz. they are nearly all from 

 three inches to five inches long, and little more than half an inch 

 thick (see Fig. 2, natural size). They are by no means constantly 

 hexagonal, but vary from three- to seven-sided. In all, however, 

 the faces are clean and well defined, and the angles sharp. The 

 rock material of which they are composed is a brittle sandstone, so 

 hard that the columns clink when struck against one another, but 

 loose and porous in appearance when examined on a broken surface. 

 On such fractures it is seen to be made up of small irregular 

 quartzose grains, which look as though fused together at their edges, 

 but with vacant spaces as great as themselves left abundantly 

 amongst them. Towards their exterior the columns are moi*e 

 compact, and upon the joint surfaces themselves the quartz grains 

 are more or less fused together ; indeed, we may often see here a 

 continuous layer of vitreous quartz over a small area, where the 

 outlines of the component grains are only indicated, if at all, by 

 such a granular appearance as is seen in a compact quartzite. It 

 is as though each prism had been half fused in a hot mould, so 

 that the clean surfaces and sharp angles were most altered by the 

 heat, and the interior was less so. 



There is no such special structure on the joint surfaces of the 

 ordinary Quader Sandstein. Comparing the fractured surface of one 

 of these columns with rock specimens from the ordinary Quader 

 Sandstein, I cannot detect any very striking differences between 

 them. The columnar rock is even of looser build, internally, than 

 usual ; but the whole column is harder and less easily crumbled 

 than the unaltered rock, this being especially the case with the 

 surface layer. 



A microscopic examination and comparison of prepared thin 

 sections of the two kinds of rock, the altered prismatic and the 

 ordinary sandstone, does not yield us any special characteristics of 

 distinction. Professor Bonney, who has kindly examined the slides, 

 writes concerning the columns, that they " consist of angular to rather 

 rounded quartz grains, with one or two which may be decomposed 

 felspar, cemented by a rather abundant dark substance (quaere if 

 this be not in part discolouration from powder used in preparing the 

 slide, as the texture is rather open). In the quartz many minute 

 inclosures, almost always less than "001", appearing when highly 

 magnified of a pale purplish colour, and in form rather like irregular 

 tubes. In one or two there was an appearance which might indicate 

 a very minute bubble, but in most cases there was no sign of this. 

 Very little difference between the two slides, except that the 

 cementing substance in that from the column is paler, and the quartz 

 grains look a little cracked." 



We may next notice the well-marked layers in which the prisms 

 are arranged. This is in no way connected with stratification lines 

 of deposition, but is a superinduced joint-structure such as occurs 

 not unfrequently in prismatic basalt masses, e.g. in the great basalt 



