104 



prisms are mostly quadrangular, and the unavticulated 

 pillars of Ballylagan, have, likewise, the same property, 

 in an inferior degree; while the basalts of our whyn dykes 

 have often their subordinate prisms finished with great 

 neatness.* 



But the forms, into which our basaltic masses divide, 

 are, by no means, limited to prismatic alone. The pyra- 

 mid is a common tigiu'e in our v/hyn dykes; and the 

 most perfect joints, of the Giant's Causeway pillars, par- 

 take both of the prism and of the pyramid, and have 

 also a mixture of curve, and plain surfaces : the latter 

 iu number equal to the denominator of the figure; while 

 the former amounts to double that number, plus two. 

 Thus, a pentagon joint, taken from one of our most 

 perfect pillars, has five plain, and twelve curve surfaces; 

 but curve surfaces are irreconcileable, either to crystal- 

 lization or desiccatio7i.'\- 



We have another vaiiety of basalt, %vhGse surfaces ex- 

 ternal, 



* This subordinate construction is well illustrated, in a drawing of three 

 ■prismatic stones, taken from a great whyn dyke (now used as a quarry), 

 nearly two miles west from Belfast. 



The constituent figure here, is a triangular prism, whose angles, at the base, 

 seem double the angle at the vertex. 



My ingenious friend, Dr. M'Donnel, to whom I had mentioned the curious 

 construction of our whyn dykes, was so struck, when he' saw the prismatic 

 stones, of which tliis dyke is formed, extracted from the quarry, that he em- 

 ployed a painter, to make a drawing of some of them; and lie was so good 

 as to give me a copy. 



f The acute-angled, triangular pyramids, which ascend from each angle 

 of the joint, and often reach up to the middle of the incumbent one, have 



their 



