AND SPHEROIDAL STRUCTURE. 149 



hut on closer examination it becomes evident that this apparent 

 division is caused by a change in the direction of the faces of the 

 column, so that for a very short space, at this point, they are no 

 longer vertical, but bend very slightly in or out. Nearly all these 

 apparent bands curve a little upwards. Sometimes they correspond 

 with actual cross joints; at others there may be very fine con- 

 cealed cross joints; but in many there is not the slightest sign 

 of any division at all. Still these bands indicate a tendency to 

 division ; for at a short distance the columns become less regular, and 

 the real cross joints (still curving upwards) more numerous, so that 

 the lava is first divided into irregular rectangular blocks, with 

 curved upper and under surfaces, and then broken by irregular 

 curving joints, which in a few cases approach the curvitabular 

 structure already described ; so that in one part the columnar, in 

 another the curvitabular structure is dominant. 



To proceed to the spheroidal structure. Professor J. Thomson 

 regards it as the result of a process of exfoliation due to the action 

 of the weather on a tolerably regular-shaped homogeneous mass. 

 Though undoubtedly cuboidal blocks of rock have, for obvious 

 reasons, a tendency to weather into rough spheroids, and though 

 further action of the weather might occasionally produce concentric 

 exfoliation in such spheroids, yet I trust to show that this explana- 

 tion is wholly inadequate in the present case, although doubtless 

 the weather has great effect in developing the structure. The ex- 

 amples which I am about to mention will, I think, establish these 

 two propositions : — 



(1) Spheroidal structure is to be seen in rocks which are not 

 homogeneous and are not at all cuboidal in form. 



(2) Spheroids may be found in columnar rock which have evi- 

 dently formed inside a prism, the exterior of which was not broken 

 by joints. 



Spheroidal structure has been observed in plaster on a wall*. 

 A very fine example of it in bedded shale is figured by Mr. Jukes in 

 his * Manual of Geology 'f. I have seen it well developed (of an 

 ellipsoidal form) in a lenticular fragment of shale caught up in 

 basalt on the Fifeshire coast, near Elic. A very fine instance of it 

 may be seen in volcanic ash near the village of Santa Lucia (Yalle do 

 Cordevole, Italian Tyrol). Here it is so conspicuous that the rock 

 at a short distance might be readily mistaken for a decomposing 

 basalt. Instances of it can also be found in the agglomeratic ash 

 of the Binns, Burntisland (Fife). The annexed diagram (fig. 10) 

 will show that here the structure is wholly independent of the form 

 of the rock. Sometimes also it occurs where a distinct stratification 

 may be observed, through the planes of which it cuts. 



Again, perlitic J obsidian and pitchstone is a true case of sphe- 



* Geol. Mag. vol. viii. p. 333. t Manual of Geology, p. 311 (ed. Geikie). 



| I distinguish, of course, perlitic from sph;ernlitic structure. The best spe- 

 cimens which have come under my own notice have been from Hungary : for 

 the opportunity of examining a very interesting example I am indebted to Mr. 

 Judd'e kindness. 



