313 
duced to little more than the thickness of half-a-crown. Little 
margaceous nodules, the mineral bezoars of some, composed of 
various laminae of soft earth, of different colours, may be frequently 
found in the gravel-pits in England, and in large heaps in several 
parts of Italy ; appearing to differ from silicious pebbles, only in 
not having been impregnated with silicious matter. That such an 
aggregation of particles of clay and sand, as may form a fit sub- 
stratum of pebbles, may even now take place, and that these bodies 
may attain the forms of pebbles by attrition and the action of water, 
is very probable. In what manner that process, on which their 
induration depends, is accomplished, is not, however, easy to 
ascertain. 
The opinion of Mons. Reaumur, as to the manner in which this 
part of the operation, in the formation of pebbles, is accomplished, 
appears to be not only perfectly consonant with the appearances 
which pebbles yield ; but also with those which are yielded by the 
several silicious petrifactions, which are more particularly the ob- 
ject of our inquiry*. He remarks, that, by a coarse operation, 
emery is reduced to powder, and suspended in water, for several 
days; but he observes. Nature may go much further; for the par- 
ticles which water detaches from hard stones, by simple attrition, are 
of an almost inconceivable degree of fineness. Water, thus impreg- 
nated, he supposes, contributes to the formation of pebbles, by 
petrifying of stone, as it were a second time : stones, already formed, 
but remaining of a spongy nature, acquiring a flinty hardness by an 
impregnation with this crystalline fluid. Tims impregnated, masses 
of different clays, chalks, marles, and boles, become flinty stones; 
and thus, he remarks, this crystalline fluid, filtering through a porous 
lime-stone, will deposit its very minute particles, in all the inter- 
stices, existing between the particles of the stone, which at last they 
* M6moires de l’Acad6niie des Sciences, &c. 1721, p. 251. 
S S 
VOL. I. 
