IN ROCKS OF DISSIMILAR ORIGIN. 339 
The occurrence of bands formed by the coalescence of spherules. 
The presence of glassy matter in strings or bands is, however, 
one of the safest tests. 
The rounded contours of porphyritically imbedded crystals also 
serve to distinguish a once-vitreous lava from a felstone. 
There is a close relation in chemical composition between hyaline 
rhyolites, felsites, &e. 
The devitrification of hyaline rhyolites may also result in hiille- 
-flinta or felsite without any marked change in the chemical composi- 
tion of the rock; and this indicates that devitrification is rather to 
be regarded as a physical change. 
Deductions. 
In this paper I have endeavoured to show:—(1) That many 
rocks to which different names have been applied are identical, some- 
times in mineral constitution, sometimes in structure, and that they 
merely differ in the relative dimensions of their constituent crystals, 
grains, or fragments. 
(2) That the rounded forms of crystals and fragments may be due 
to various causes already specified. 
(3) That vesicular and amygdaloidal structure may be due to 
various causes already specified. 
(4) That other spherical or spheroidal structures occurring in 
certain rocks are due to various causes already specified. 
(5) That in view of the foregoing considerations it may be well to 
roughly classify all rocks as eruptive and clastic, and to subdivide 
them according to certain structural peculiarities, as indicated in the 
classification proposed in this paper. 
(6G) That both angular and rounded fragments frequently occur in 
the same rock, and consequently that sedimentary rocks cannot be 
characterized as consisting exclusively or especially of rounded 
rains. 
(7) That the microscopic characters of felsitic matter in some cases 
may, in others may not, afford a clue to the origin of certain rocks ; 
and that, since, in many cases, lavas and volcanic ejectamenta are 
often bedded, just like sediments, no distinction can be found, either 
by observation in the field, by pyrognostic deportment, or by micro- 
scopic examination, between certain devitrified lavas, felstones, and 
arenaceous sediments, when they all present felsitic characters, and 
between certain grits, breccias, tuffs and volcanic ashes. 
Let us for a few minutes consider what are the bearings which 
recent petrological investigations have with regard to past geological 
work. With every modern appliance it is often difficult or im- 
possible to pronounce more definite opinions upon the character or 
origin of certain rocks than those which were enunciated years ago. 
The more we learn about some rocks the less are we disposed to 
assien them definitely to any particular origin; and we are not un- 
frequently compelled to fall back upon the vague terms of our pre- 
decessors as harbours of refuge in which to shelter until more 
