V 



Sect. YL] 



GEOLOGY. 



181 



from those produced by ancient glaciers is, at present, 

 reat desideratum in geology. M. Agassis' work on 



Glaciers, with its admirable plates, ought to be procured 

 by any one going to the colder regions of the north or 



south. 



Erratic boulders occur in Europe, N, America, and in 



the southern parts of S. America, which, it is believed by 



most geologists, were transported by ice ; those near 



mountains, by ancient glaciers ; and those on the low- 



lands, by floating ice. 



Erratic 



boulders, when not of 



gigantic size, may be confounded with rounded stones, 



tne coast- 



transported by occasional great floods or 



action of the surf during slow changes of level of the 



land. Masses of granite, from often disintegrating into 



large, apparently water-worn boulders, and then rollin 

 downw^ards, have several times been erroneously described 

 as belonging to the erratic class. Where the nature of 

 all the rocks in the vicinity is not perfectly known, great 

 size and the angularity of the fragments (though by no 

 means a constant concomitant) are the most obvious dis- 

 tinctive characters; but even wdien the surrounding 

 country is not at all known, the composition of a single 

 isolated hill or small island may easily be ascertained, 

 and if large fragments of foreign rock lie strewed on its 

 surface, these may be assumed almost certainly to be 

 erratic boulders. Here, how'cver, a caution has been 

 found necessai y ; for in the case of fragments of sedirneu- 

 tari/ rocks, they may be the last remnant of a denuded 

 overlying formation. Wherever erratic boulders are 

 found, their composition, form — especially attending to 

 whether they are angular, water- worn, or scored, an<i 



\ 



