258 ON THE DESTRUCTION OF ROCKS. niately resolved into loose earth. But these changes are not limited to prismatic trap only, since they occur in other and different forms ; being attended with the same final result, whatever the shape of the block may have been, namely, tbe ultimate production of spheroids. Although, in some of these instances, it might be imagined that the effect of desquamation was pro duced, as in the case of the artificial granite columns, by the exposure of the surfaces to the air, it will, I believe, be found that, in nearly all, they truly depend on an internal concretionary structure. The chief argument for this opinion will be found to consist in tbe effects which take place in those traps that are not jointed, as well as in those masses which affect a prismatic fracture without being absolutely divided into prismatic forms. In these, as in some of the cases of exfoliating granite, the desquamation is of a complicated nature, referring to more than one centre. Thus, in a single unjointed prism, the same result takes place as in those with joints; numerous spheroids being discovered in its length, resulting from the progress of desquamation. The same effect is produced in those irregular masses which are charac terized merely by a prismatic fracture ; as the exfolia tion commences in many different places, referable to different central points, so as to leave, in the same way, a number of spheroidal bodies imbedded in a mass of loose crusts and clay. It is further, indeed, often to be observed, that in cuboidal or otherwise irregular blocks which have neither prismatic form nor ten dency, there are several centres of exfoliation: nu merous balls being thus finally extricated from a single solid block, of which all the surfaces are equally- exposed to the action or contact of the atmosphere.