330 F, RUTLEY ON COMMUNITY OF STRUCTURE 
leave definitely shaped cavities or moulds of the departed crystals ; 
but these in time become further weathered into more or less sphe- 
rical cavities, so that at length the rock resembles a scoriaceous lava. 
The rock at Washfield, near Nether Stowey, is a good example of 
this. 
We must not, indeed, always impute an eruptive origin to every 
rock which exhibits traces of vesicular structure, since vesicles may 
even be met with in sediments. They are, however, very rarely seen. 
The marsh-gas derived from decaying animal and vegetable matter 
may account for their presence in the mud of ponds, brooks, &c. ; 
for, if we stir such mud gently with a stick, bubbles of gas are freely 
disengaged, and it is consequently evident that they must exist, 
pent up within the mud. 
If the decomposing matter lie beneath merely a thin covering of 
mud the gas-bubbles, as they become naturally disengaged, will 
probably pass completely through it; but, in other cases, where the 
superstratum of mud is thick, the bubbles may only pass some little 
distance into the mud; and if such a deposit became quickly dry 
(as it might do on a muddy shore or in an estuary when the tide 
went out), it is quite possible that the gas-bubbles would form per- 
manent spherical cavities. Under similar conditions raindrops, 
worm-tracks, and footprints have formed permanent records, and 
like conditions ought to suffice for the preservation of a vesicular 
structure in fine muddy sediments in which organisms have decom- 
osed. 
A That records of this kind are of rare occurrence is doubtless due 
to the obliterating disturbance produced by the returning tide. 
That they are so seldom observed is also due, in part, to the infil- 
tration of mineral matter in solution after the consolidation of the 
deposits. 
In rocks of this kind which have undergone little or no compres- 
sion such cavities would be spherical; but in all old deposits which 
haye undergone the pressure of a considerable thickness of superin- 
cumbent strata, we should naturally expect to find them ovate or 
lenticular. The irregular form of the vesicles so commen in eruptive 
rocks must be attributed to forcible injection of gas or vapour, com- 
pression, or the irregular motion of surrounding matter in a viscid 
condition or sometimes even in a state of ebullition. A pisolitic 
structure is occasionally met with in fine pumice tuffs, and is stated by 
Cotta* to be due to the action of raindrops, being a phenomenon 
similar to that which is produced at the present day when showers 
of rain accompany showers of volcanic ashes. The spherical form of 
volcanic bombs is due to the rotation of masses of molten matter 
during their projection and fall through the air. We have now 
reviewed the principal causes which account for the rounding of 
erystals and fragments of mineral matter and for the spherical or 
spheroidal cavities and kernels which occur in rocks; but there are 
other spherical and spheroidal forms which have yet to be considered, 
and which are also of considerable petrological interest. In many 
* Rocks Classified and Described (London, 1866), p. 309. 
