According to Mr: Gregory Watt the columnar structure of Basalt is due to the .pressure of numerous spheres or spheroids on each other during the cool ing of the rock, such spheres or sphe roids being produced in planes of refri geration or absorption. This author took seven cuts. of an amorphous Basalt named Rowley Rag, kept it in fusion for more than six hours, and cooled it so gradually that eight days elapsed before it was taken from the furnace. The shape of the mass was uneven, and while the thinner portion was, in consequence of more rapid cooling, vitreous, the thicker was stony, the one state passing into the other. It was observed that numerous spheroids had been formed, sometimes two inches in diameter. They were radiated with distinct fibres, the latter also forming concentric coats when circumstances were favourable to such an arrangement. When the temperature had been sufficiently continued, the centres of the spheroids became compact before they attained the diameter of half an inch. When " two spheroids came into contact no penetration ensued, but the two bodies became mutually compressed and sepa rated by a plane, well defined, and invested with a rusty colour," and when several met they formed prisms. • The following are Mr. Gregory Watt's inferences from these facts : " In a stratum composed of an indefinite number in superficial extent, but only one in height, of impenetrable spheroids, with nearly equi distant centres, if their peripheries should come in contact in the same plane, it seems obvious that their mutual action would form them into hexagons ; and if these were resisted below, and there was no opposing cause above them, it seems equally clear that they would extend their dimensions upwards, and thus form hexagonal prisms, whose length might be indefinitely greater than their diameters. The farther the extremities of the radii were removed from the centre, the greater would be their approach to parallelism ; and the structure would be finally propagated by nearly parallel fibres, still keeping within the limits of the hexagonal prism with which their incipient formation commenced ; and the prisms might thus shoot to an indefinite length into the undisturbed central 'mass of the fluid, till their structure was deranged by the superior influence of a counteracting cause." (' Observations on Basalt, &c. ;" Phil. Trans.,' 1804.) According to this theory, which is certainly the best hitherto framed to account for the columnar structure of Basalt, the irregu larity of the prisms would obviously depend upon the un equal -distances of the centres of the spheroids, and the con sequent unequal pressure ; and it is further inferred that the joints sometimes observable in basaltic columns correspond with the concentric coats noticed above. Two of the
most beautiful examples of columnar Basalt hitherto dis covered are found in the British Islands,one forming the Giant's Causeway, on the north coast of Ireland, and the other at Staffa, among the Hebrides. The largest columns yet ob served are found at Fairhead at the former place, where, according to the accurate measurement of some by the Ordnance Trigonometrical Sur vey of Ireland, they are 317 feet in height, the sides of these enormous prisms occa sionally measuring 5 feet.
Some non-columnar Basalts present no trace of any particular arrangement of parts, while others show a globular structure, so that when the rock becomes more decomposed it has the appearance of numerous bombshells and cannon-balls cemented together by a ferru ginous substance. This globular structure is sometimes also apparent when the decomposition of the rock has not been considerable, being well exhibited in the concentric arrangement of coats of Basalt round centres at variable distances from each other.
Other Basalts are amygdaloidal, containing a variety of substances, such as Agates, Onyxes, and other minerals, which have been infil trated into cavities formed by bubbles of gas or vapour while the rock was in a state of fusion. As these bubbles have sometimes been lengthened by the flow of the rock before it finally cooled, the infil trated contents filling such lengthened cavities have the appearance of almonds sticking in the mass of the rock, whence the name amygda loid. When, as sometimes occurs, a great tabular mass of Basalt is composed of superimposed beds, some columnar, some amorphous, and others anfygdaloidal, these characters are sufficient to authorise a conclusion that the whole nasa has not been produced at one upburst of Basalt, but that there were several flows of melted matter to which different conditions gave different characters ; the atuygdaloidal struc ture particularly pointing to the absence of very considerable premium upon the Basalt so ehameterised, before it became solid.