ORIGIN OF COLUMNAR STRUCTURE—JAUJAC—STAFFA. 13 
emission (the Coupe de Jaujac, near the village of that name). Near Jaujac, 
where the cliff may be, perhaps, from 60 to 70 feet high, only one-fourth of the 
height is occupied by the columnar range; down the course of the valley it rises 
to one-third and one-half of the height of the cliff (which is also increasing), it 
then divides into two columnar beds, with an amorphous portion between, and 
finally presents the magnificent display of unbroken columns in almost its entire 
height, already referred to. 
We have seen, that, sharp asthe separation appears of the columns from the 
amorphous or slightly-columnar part, it is impossible to deny that both are the 
production of a single eruption. The cause of the abruptness of the separation 
appears, in the present state of our knowledge, altogether inexplicable. Mr 
Scrors, indeed, gives the following elucidation, but I doubt whether it can be 
considered as, in any degree, satisfactory. ‘“|This| singular difference of struc- 
ture may be accounted for, I conceive, by supposing the elastic vapour, contained 
in the lava, to have escaped from the upper part, through fissures of retreat, 
formed in irregular directions as the mass cooled; while the immense pressure, 
acting on the lower part, would, at the same time, prevent the escape of the con- 
tained vapour, and effect its condensation, thereby allowing the mass to consoli- 
date in a regular and tranquil manner, such as would facilitate the establishment 
of straight and vertical axes of contraction, and the formation of very regular 
hexahedral columnar concretions.”* Whatever be the cause, it would appear to 
become more energetic after the lava has flowed to a considerable distance ; and 
from the regularity with which the limit of the columns follows the sinuosities of 
the base (see Plate VI., fig. 3), it would seem to depend more upon the rapidity 
of the cooling action below than upon the superincumbent load of lava. 
It is a very curious circumstance that this sharp limit of columnar forma- 
tions occurs also in circumstances which would appear very different from the 
volcanic flood of Jaujac, which undoubtedly occurred in the open air exactly like 
any modern eruption. A recent visit to the Island of Staffa has satisfied me that 
the basaltic bed in which Fingal’s Cave occurs, is constituted precisely like the 
lavas of the Vivarais, and that the sharp cessation of the columns, and the abrupt 
transition to a nearly shapeless bed of basalt, is not due to a superposition of 
other matter, but simply to the limited sphere of action of the crystallizing or 
concretionary forces; we have there, as at Jaujac, perfect evidence of the con- 
temporaneous formation of both beds; and, indeed, the apparent section of the two 
would be perfectly identical, the compact part commonly projecting over the colum- 
nar part, asin Plate II., fig.6. Dr MacCuttocu, who makes no allusion to the con- 
temporaneity of the beds, or to the cause of their distinction of character, minutely 
specifies the exact perpendicularity of the principal range of the columns of Staffa 
* Geology of Central France, p. 152. 
VOL. XX. PART I. D 
