Mr Poulett Scrope's Considerations on Volcanos. 347 



freely as the mass below contracts, no fissure, or very few will be formed 

 parallel to that surface ; and by the inward propagation of the retreat, the 

 hexagons will be lengthened into hexagonal prisms. The slower the pro- 

 cess of solidification, and the finer the grain of the lava, the more regular 

 will be the prisms, ceteris paribus ; hence the interior, or lowest parts of a 

 current alone, in general, require this structure, which does not become 

 visible till denudation has exposed these parts. This is rarely the case 

 with recent lavas ; and hence arises, according to our author, the common 

 error of supposing the columnar division confined to the older basalts. 

 This structure is very frequent in dikes, both in the older and recent vo- 

 canic formations ; the columns being always perpendicular to the sides of 

 the dike. When lava rests on a convex surface, the columns diverge ; 

 when on a concave, converge upwards ; being always perpendicular to the 

 surface on which the process first acts. The author remarks, that those of 

 the peaks of basalt, which are so numerous in basaltic districts, will be 

 found to consist of a group of convergent columns; this disposition afford- 

 ing the maximum of resistance to the action of rain and frost, in separat- 

 ing the columns, and breaking up the bed, of which the remainder has 

 probably been destroyed in this manner. More than one kind of division- 

 ary structure may occur in the same rock ; smaller prisms are sometimes 

 formed within the large. The globiform structure is next accounted for, 

 and its occasional subdivision into radiating prisms, or concentric leaves. 

 The angulo-globular structure accompanies a tendency to the formation of 

 globular concretions. The tabular, lamellar, and slaty, or schistose divi- 

 sionary structures, are supposed to be confined to lavas in which the crys- 

 talline particles are disposed more or less conformably, owing to which 

 their mobility is considerable in the direction of their parallel plane sur- 

 faces, and null in the transverse direction. Hence, retreat fissures will be 

 produced in abundance, parallel to the largest plane surfaces of the crys- 

 tals, and few or none will be formed transverse to these. By the fre- 

 quency of such transverse fissures, the cubical or rhomboidal structure is 

 produced. 



The author next touches on the question, as to the cause of the differ- 

 ence of mineral composition in lavas. He inclines to attribute this variety 

 to certain alterations undergone by the rock, originally of an uniform 

 (perhaps granitic) composition, during its rise to the surface of the globe ; 

 which was probably attended by repeated alternations of intumescence and 

 reconsolidation, from changes in the relative proportions of the intense 

 heat and pressure to which it was subjected. The principal varieties of 

 lava are found in nature to have been usually produced successively, often 

 alternately, from the same or proximate vents. The opinion of the anta- 

 gonism of trachyte and basalt, put forth by Humboldt and Beudant, is 

 combated by our author, and numerous examples adduced of their suc- 

 cessive emission from the same volcano. He then notices on the error of 

 limiting the production of trachyte or basalt to particular ages of the globe, 

 or making " formations" of them — both are produced before our eyes by 

 recent and still active volcanos ; the error arises from the terms trachyte 



4 



