ON THE STATISTICS OF DUKHUN. 223 
angles disintegrate and a slope results. But most usually 
three or four beds of amygdaloid are found between two strata 
of compact basalt ; the former disintegrates, leaving a slope, 
which is not unfrequently covered with forest trees, forming a 
picturesque belt. ‘The basaltic scarp remains entire, or it 
may be partially buried by the debris from the amygdaloids 
above ; but its great thickness usually preserves it from obli- 
teration, and it rises from the wood below with majestic 
effect, its black front being finely contrasted with the rich and 
lively green of its sylvan associate. It is these strata, arranged 
in slopes and scarps, repeated three or four times, and 
so commonly met with in insulated and other mountains in 
Dukhun, that constitute the amazing strength of the hill 
forts of the country, leaving a succession of natural walls 
encircling a mountain. This feature did not escape the ob- 
servation of Captain Dangerfield in Malwa, who says, ‘‘ From 
the great difference in the resistance made to decomposition 
by these trap and amygdaloid beds, their exposed ends acquire 
a very distinct degree of inclination and character; the amyg- 
daloid forming a great slope and affording a loose mould 
covered with vegetation, the trap retaining its original per- 
pendicularity and dark bareness.” 
In the alternation of the strata there does not appear to be 
any uniformity ; but the general level, thickness, and extent of a 
stratum are preserved, as in sedimentary rocks, on both sides of 
a valley; the basalt and hardest amygdaloids being traceable 
for miles in the parallel spurs or ranges; but the imbedded 
_ minerals, and even the texture, vary in very short distances. 
Columnar Basalt.—A great geological feature of Dukhun is 
a the occurrence of columnar basalt. The basalts and hardest 
_ amygdaloids run so much into each other that the line of sepa- 
ration is not always readily distinguishable, excepting, of course, 
the lines of horizontal stratification. I observed the prismatic 
disposition more marked and perfect in the basalt strata than 
in the amygdaloids, and the more or less perfect development 
of determinate forms was dependent on the compactness and 
limited constituents of the rocks. Basalts and amygdaloids, 
however compact, with many imbedded matters, rarely formed 
columns. Perfect columns were generally small, of four, five, 
or six sides; but the prismatic structure sometimes manifested 
itself in basaltic and amygdaloidal columns many feet in 
diameter. A bare mention of the places where they occur 
_ would testify to their extended localities, but these are too nu- 
merous for insertion in this report. 
Schistose Structure.—Following the preceding formation, I 
