274 
PEOFESSOE TYNDALL’S OBSEETATIOKS ON THE MB DE GLACE. 
Let 
rig. 5. 
1. . ...... .1 “ 4 “" rii ..... ■ ' 
whole square moves downwards with * 8 ’ • . 1 4-75 inches in 
onickest The point « moves 10 inches, the pomt i. 14 lo mches m 
Iventy-four hours, the difterential motion thus amountmg to ^ m* m 
five hLrs. Let aVcH be the shape of the figure after five hours mo- 
tion the distance being 1 hich ; then the line would be 
pxf ended to aV , and the line cd to cd'. 
But the extension of the.se lines does not mark the ^ 
to which the ice is subjected. Mr. Hopxms has shown this to 
take nlace along the line ad, which encloses an angle of 45 with the 
side of the glacier. In five hours, then, this line, if capable of )relding, 
would be stretched to . rmon • n 7i/7' 9981 
In the right-angled triangle aid' we have =2280 inches M—-8L , 
are the consequen . S yielding we do not 
proves what ice camsot do . what can J ,^^11, to show that the 
know. There is no experiment on record, which is 
substance possesses, in any measuiable degre , P 
X^el^Te— accustomed to glacier life observes, from a safe distance the pto- 
fou^fissures by which the ice is intersected, “^lere 
one of these chasms should suddenly open beneaUi toe — ^ 
ever, no fear of this. The crevasses, when &st formed, are exceea ^5 
* It may, however, be urged Extend^II ekstic ^Tto tirpoiiit of breaking and a smaU 
had been stretched before it arrived there. extension would be no measure of the extensibibty 
additional force would break it ; but is a tr s viscous mass to accommodate itself to the 
o, the string. To this I reply, that tt .s the ™ I « shto he iu e shite of 
forces which act upon it, so that in each new p r- to recover from. The idea that a glacier 
equilibriun.. If sueh a m.u.s be hrohen .t wjl by the ablest aavoc-ates of the 
« typified by sueh a string “ “ " J; i,^„ihote of Dr. Wnr«LL, in the 26tb volume of 
viscous theory ; m proof of where the lateral yielding produced by theprrsswe 
the Philosophical Magasme, p^lge 172. Cases y 
along he, fig. 5, may satisfy the s(r«» along ad ; in such a case no . 
