
Partil] | PRISMATIC STRUCTURE. 507 
columnar structure has been produced. The experiments of Mr. 
Gregory Watt were supposed to explain it by the production of a 
number of spherical concretions in the cooling mass, and the gradual 
pressure of those soft balls into hexagonal columns, as the mass con- 
tracted in cooling. He melted a mass of basalt, and on allowing it 
to cool observed that, when a small portion was quickly chilled, it 
took the form of a kind of slag-like glass, not differing much in 
appearance from obsidian; a larger mass, more slowly cooled, 
returned to a stony state. He remarked, that during this process 
small globules make their appearance, which increase in size by the 
successive formation of external concentric coats, like those of an 
onion. And he supposed that, as each spheroid must be touched by 
six others, the whole, if exposed to the same pressure acting in every 
direction, must be squeezed into a series of hexagons. To account, 
however, for a long column of basalt, we should have to imagine a 
pile of balls standing exactly centrically one upon the other, an 
arrangement which seems hardly possible. The prismatic structure 
is a species of jointing, due probably to the contraction of the rock as 
a whole, and not to the production of any internal peculiarities of 
texture. The concretionary structure associated with the columnar 
reveals a common tendency to weather out into nodular forms, 
and may be observed even where the rock is not columnar. As 
already stated, prismatic forms have been superinduced upon rocks by 
a high temperature and subsequent cooling, as where coal and sand- 
stone have been invaded by basalt. ‘They may likewise be observed 
to arise during the consolidation of a substance from aqueous solution. 
In starch, for example, the columnar structure may be well developed, 
and not infrequently radiates from certain centres, as in basalt and 
other igneous rocks. 
Mr. Mallet has investigated this subject, and concludes that “all 
the salient phenomena of the prismatic and jointed structure of 
basalt can be accounted for upon the admitted laws of cooling, and 
contraction thereby, of melted rock possessing the known properties 
of basalt, the essential conditions bemg a very general homogeneity 
in the mass cooling, and that the cooling shall take place slowly, 
principally from one or more of its surfaces.”* In the more perfectly 
columnar basalts the columns are sometimes articulated, each prism 
being separable into vertebre, with a cup and ball socket at each 
- articulation (Figs. 224 and 225). This peculiarity is traced by Mr. 
Mallet to the contraction of each prism in its length and in its 
diameter, and to the consequent production of transverse joints, which, 
as the resultant of the two contracting strains, are oblique to the sides 
of the prism, but, as the obliquity lessens towards the centre, assume 
necessarily, when perfect, a cup-shape, the convex surface pointing in 
the same direction as that in which the prism has grown. ‘This 
explanation, however, will hardly account for cases, which are not 
uncommon, where the convexity points the other way, or where it is 
1 Proe. Roy. Soc. January, 1878. 
