O'Reilly — On a Cylindrical Mass of Basalt. 239 



It is to be remarked, that of the two examples thus described, the 

 one is a vertical form in a trachyte rock, whilst the other is a hori- 

 zontal form in a basalt ; moreover, its direction is connected with 

 that of the basalt dykes of the district. 



M. Eegnaiilt attributes the cause of the forms to alterations of 

 the rock mass, with accompanying exfoliation, but does not exa- 

 mine the question whether the rock was or was not originally homo- 

 geneous, and how far the absence of homogeneity in the basaltic and 

 trachytic masses, and the consequent different rates of cooling and 

 solidification which must have depended on that absence of homo- 

 geneity, induced subsequent jointing along certain lines. 



]^ow, no mass of molten or fluid matter when in movement, and 

 in contact with bodies differing from it in temperature and in com- 

 position, can remain quite homogeneous or have all its parts equally 

 warm, and therefore, so long as it remains fluid and in movement, 

 there must be a tendency to the formation of a more or less regularly 

 banded structure, which would guide and even assist subsequent alte- 

 ration of the rock, particularly by the action of water. 



It is upon these considerations that I propose to base an explana- 

 tion of the cylindrical form in question. 



When columnar basalt is carefully examined on the cross section, 

 lines or bands of structure may be frequently observed, generally 

 parallel to the sides of the block, but not unfrequently whorl-shaped 

 or wavy. This structure may be mainly clue to hydration, but it may 

 also be connected with the original fluid state of the mass, and with 

 the relative rates of solidification of the different parts. We have as 

 an example the flow-lines in large masses of cut glass. 



As regards the basalt of the north of Ireland, we have every reason 

 to assume that they came up to surface in the joints resulting from 

 previous dislocation, and thence spread into and on the adjacent rocks. 

 Those joints traverse rocks of various constitutions, representing lavic, 

 crystalline, and sedimentary series. These rocks vary amongst them- 

 selves, as regards thickness, constitution, and states of aggregation. 

 They were differently affected by the force having produced the joint ; 

 and it is but reasonable to assume that the joint varied in outline ac- 

 cording to the nature of the rocks traversed. 



When, therefore, the basaltic dyke mass was fluid and in movement, 

 it was continuously in contact with those containing rocks, and was 

 acted upon by them in three ways : — 



1 . By the cooling of the basaltic fluid, from the sides towards the 

 central axis, and the consequent tendency to banded structure thus 

 induced. 



2. By the mechanical derangement of the conditions of regular 

 flow, wherever an enlargement or a contraction occurred in the sec- 

 tion of the joint, and by the retardation of the flow along the walls. 



3. By the chemical reactions taking place between the fluid mass 

 and such beds as were capable of being acted upon. 



