1842.] PRECISION ATTAINABLE BY OBSERVATION. 11 



all its mass, the velocity of the inferior part ought to be greatest. 

 Of course, I do not now advert to the many causes which might 

 accidentally invert this law, and which would require to be fully 

 taken into account ; still less do I mean to say that anything I 

 have now to state can be considered as critically decisive between 

 rival theories ; but my experiments certainly do show that the 

 kind of precision which I desired to see introduced into reason- 

 ings about this subject, is practically attainable, even in a far 

 higher degree than I expected. 



For example : The motion of glaciers by the measurement 

 of the distance of blocks upon its surface from a fixed point, 

 from one year to another, has marked indubitably the annual 

 progress of the ice. I do not know that any one has attempted 

 to perform the measurement in a manner which could lead to 

 any certain conclusion respecting the motion of the ice at one 

 season compared with another, or from month to month ; still 

 less has any one been able to state, with precision) whether the 

 glacier moves by starts and irregularly (as we should certainly 

 expect on the sliding theory), or uniformly and evenly ; and 

 if so, whether it moves only at one part of the twenty-four 

 hours, and stands still during the remainder (as we should expect 

 on the dilatation theory, as commonly expounded). Now, I have 

 already been able 



1st. To show and measure the glacier motion, not only from 

 day to day, but from hour to hour ; so that I can tell nearly 

 what o'clock it is by the glacier index. That you may have 

 an idea of the coincidence which these experiments present, I 

 give you the longitudinal motion of a point on the Mer de 

 Glace during four consecutive days. 



15.2 inches. 16.3 inches. 17.5 inches. 17.4 inches. 



2d. This motion, evidently incompatible with sudden starts, 

 takes place in the glacier as a whole, undisturbed by the most 

 enormous dislocations of its surface, for these measures were 

 taken where the glacier was excessively crevassed. 



3d. This motion goes on day and night, and if not with 



