298 PEOFESSOE TYNDALL ON THE \H;LNED STEDCTLLRE OF GLACTEES. 
%. 10, represents a piece of ice detached from the gaUery of inatration ; all the bubbles 
are gi'eatly flattened. But what is most extraordinary is, that far ftom bemg mirform 
ftattming is different in each fragmsnt ■, so that the bubbles, according to the 
they ofter, appeal- either very broad or very thin. I know of no more significant fact t an 
this, since it deimstrates that each fragment of ice is capalie of undergoing in the inte, lo, 
of the olaeier a nroper displewement indqiendenthj of the movement of the whole. 
“The same flattening of the bubbles,” continues M. Agassiz, “is found at a greatei 
depth. While engaged in my boring experiments, I observed attentively the agments 
of ice brought up by the borer. I found in them almost flat bubbles, perfectly sim ar 
to those of the fragment figured above, at all depths from 10 to 6o^ metres. It follows 
hence that a strong pressure is exercised on the interior of the glacier. 
The description of the “ flattening” here given is correct : all obseiwers agree m corro- 
borating it, and eveiy observer with whom I am acquainted draws substantiaUy the same 
conclusion from the phenomenon that M. Agassiz does. Professor Thojisox s specula- 
tion upon the subject is pai-ticularly refined and ingenious. 
Mr Josh Ball converts the flattening of the bubbles into endence agamst the 
pressure theory of the structure in the following way:~“As Agassiz has pomted out 
writes Mr. Ball, “and I have frequently verified his observations upon this pom . 
though the air-cavities show traces of compression reducing them to the form ot flat- 
tened lenses, the directions in which they are flattened are most various, and show no 
ccmstant relationship to the planes of the veined structure. Here then we hate diiea 
evidence that separate portions of the ice have been acted upon by pressin-e sufficient m 
amount to modify their internal arrangement, but that these pressures have not acted in 
the same, or nearly the same direction.” _ _ ^ 
Granting the inference that the observed flattening “ furnishes direct eindence o 
pressure, the foregoing argument would, I confess, be a very formidable one. t t le 
bubbles are thus flattened by pressure, and if the veined striictiu-e, as I contend, be t e 
result of pressure, and approximately at right angles to the direction of the 
ought to have the bubbles squeezed out in planes parallel to the structure. ^ le . ac 
that the bubbles are not so squeezed out, would then afford a strong iiresiimption t lat 
the structure is not produced by pressure. I expect, however, to be able to prove that 
the shape of the bubbles is not a “direct evidence” of pressure, as hitherto assumed 
and 1 think, as I do so, it will be seen how necessary it is to associate experiment uit 
an inquiry of this kind, if we would read aright our observations. 
In a paper in the Philosophical Transactions on the Physical Properties of Ice, I have 
shown that when a sunbeam traverses a mass of ice, the latter melts at innumeia i e 
points in the track of the beam, and that each portion melted assumes the form, not 
of a globule, but of a flower of six petals. The planes in which these flowers are formed 
are independent of the shape of the mass and of the direction of the beam through it ; 
they are always foimed 'parallel to the surface of freezing. 
This is a natural consequence of the in aimer in which the particles of ice aie se 
