206 
A comparison of the two analyses A and B shows a remarkable simi¬ 
larity, and indicates that the materials were of the same composition. 
The greatest difference is in Alumina 1.54 per cent ( + ) and Silica 0.82 per 
cent. ( -); otherwise the figures are singularly alike. The specific gravi¬ 
ties are identical. 
The analysis was performed with all the precautions laid down by Dr. 
Hillebrand and were complete before the Tasmanian report came under my 
notice. 
In a note discussing his 'own figures, Dr. Hillebrand says “ there is 
probably some error in the value given for one of the larger constituents 
since the summation is in excess *of 100, even without the water which in 
all probability is present /’ 
In the present analysis the same remark applies, except that the error 
is on the minus side, and would probably bring up the alumina figure as 
that is determined! by difference in the usual manner. The summation 
(99.78 iper cent.) is, however, quite within the range of safe work. 
PLATE XXXIII. 
Obverse Side. 
% 
Fig., 1.—Represents a dumb-bell form. This is considered to be the* 
‘ : bleb ” formed bv two bubbles of obsidian coalescing. The surface 
represented would be uppermost as the bubble floated, and it would there¬ 
fore be inside the bubble, which would be attached around the periphery. 
From Mackenzie Creek, Horsham. 
Fig. 2.—Represents a bleb of a single bubble which was originally 
like Fig. 3, but much larger and thicker. The thin rim which surrounds 
Fig. 3 has flaked off, leaving a stopper-like form slightly bevelled where- 
the flakes have split off. From Coolgardie, Western Australia. 
Fig. 3.;—This is a bleb resembling a button in form. The rough 
margin is where the single bubble was attached to it. In blowing ordinary 
bubbles frequently a bleb forms , at the bottom, and in this case the bleb 
is of obsidian and of the form shown. The remarkable symmetry of these 
buttons can be accounted for in this way. From Byaduk, Victoria. 
Fig.. 4. This is a hollow sphere of obsidian 2 inches in diameter- 
a bubble, in fact •—- and it is the key to the origin of the 
whole of these interesting objects. It is a bleb of a large- 
bubble, but in this case another bubble with thick walls has 
been formed as a bleb. It is easy to understand that this bubble under 
other conditions would have expanded until the walls were very thin and 
filled with some light, highly heated gas. The rim around this bubble 
shows where it was attached to the large bubble of which it formed the 
bleb. I hese obsidian bubbles were formed, not during violent eruptions, 
or they could not have escaped destruction, but during periods of com¬ 
parative quiescence, and when there was but a gentle ebullition. As t.hev 
escaped from the crater they were floated away by the air currents, and 
landed eventually in all sorts of situations, and in some cases at great dis¬ 
tances from their source. When found on the surface, the surfaces shown* 
