119 
cases this is so large that the surrounding oxide of man- 
ganese becomes very thin and a delicate tube is formed, 
while in many specimens the calibre is so small that the 
perforation can only be seen on close examination. The 
columns and threads are not all straight, but many are bent 
in various directions and not unfrequently branched — the 
axial canals in all cases being preserved throughout the 
bifurcations. 
These interesting structures in external appearance are 
rough and dull, but cross-sections exhibit a very distinct 
resinous lustre. The size of the columns is not great, their 
diameter never exceeding Jth of an inch, and their length 
being seldom much more than one inch ; the tubular threads 
are sometimes four or five inches long, but their diameter 
is very much less than that of the columns, few of them 
having a section greater than - 2 V inch across. 
To account for the formation of these columns and threads 
appears, at first sight, a somewhat difficult matter, but after 
carefully taking into consideration the position of the mine- 
ral and the nature of its surroundings, an explanation is 
afforded which has been definitely proved to be correct by 
Mr. Dale, of Macclesfield. 
The position which the greater part of the oxide of man- 
ganese occupies is about 14 feet from the surface, a depth 
to which the roots and rootlets of the surrounding wood are 
able to penetrate by means of the numerous cracks in the 
strata. Rootlets are to be seen in the fissures in many parts 
of the section, and in those which are lined with oxide of 
manganese they are especially abundant. They stretch 
across the cavities, and traverse the surface of the black 
mineral in all directions ; and after closely searching for 
some time, it is possible to find specimens to illustrate all 
stages of the conversion of these organic bodies into columns 
and threads of oxide of manganese. 
Many of the rootlets, probably in their first stages of de- 
