Correspondence — T. G. Bonney. 89 



Bidymograptus lifidus beds : blue shales -with ocasional ashes, passes down into : — 



D. hinmdo beds : dark calcareous shales, with some ashes. 



Calymene ashes, with some calcareous material and occasional shaly beds. 



D. balticus flags, quartzose flags with shaly partings. 



Easal grit. 



Unconformity. 



Shumardia shales, passing down into: — 



Asaphellus flags. 



Bellerophon beds : hard shales, with Parabolinella Salteri in upper part. 



Dietyonema band, 1 j. n j 



,7.. / , 1 ' > not well exposed. 



Niooe beds, j ^ 



Peltura beds, with Niohe and Fsilocephalus near the top, and Feltura abundant in 



the middle part. These are separated by thin beds with Orthis lenticularis from : — 



ParaboUna beds, mth P. spinulosa. 



The unconformity marks the base of the Arenig Series. The 

 Orthis limestone yields Lower Caradoc Sandstone forms, and dips 

 beneath the great Black Shale Series, which occupies the adjoining 

 belt of country. 



The" author discusses the relationship of these divisions to 

 corresponding beds of other areas. He gives a description of the 

 intrusive igneous rocks, and some account of the structure of the 

 district and the nature of its srlaciation. 



oos,ia:BSDPon:srnD.BisroE. 



CAVITIES IN CRYSTALLINE ROCKS. 

 Sir, — The cavities in the granite of Madagascar, described by the 

 Eev. E. Baron in your January number, appear to be even more 

 remarkable than those in Corsica and Portugal. But his explana- 

 tion, though very ingenious, is, I think, attended by difficulties of 

 its own. Between the ordinary drusy cavities and these large 

 smooth hollows the gap seems wide and incompletely bridged by 

 any of comparatively small size. Also, might we not expect that 

 such hollows would sometimes be cut into by quarrying, and then 

 have attracted notice ? I cannot remember to have read of their 

 occurrence, and have never met with anything of the kind, though 

 I have been in several quarries, including the one mentioned at 

 Ajaccio itself. Again, several of these hollows at that place occur 

 rather as depressions on the surface of a block than as holes in it, 

 and in positions unfavourable to the idea that any fairly thick slice 

 has been removed from the rock in comparatively recent times. 

 Of such the cavity represented on p. 390 of the Geol. Mag., 

 August, 1904, may serve as an example, and I remember others 

 in similar positions on the slope behind Ajaccio. Moreover, as the 

 granite is generally in good preservation, showing, as Mr. Baron also 

 remarks, only a little surface-decomposition, it must be weathering 

 away slowly, and it is rather surprising to find the cavities retaining 

 their superficial glaze. His description, however, seems to exclude 

 the explanation of sand blast, and makes that of water corrosion 

 still more difficult. But a serious objection to his own hypothesis 

 has, I think, been overlooked. At least one of the cavities described 



