422 SIR CHARLES LYELL 



numerable bones and skeletons of birds, quadrupeds, 

 reptiles, which tenanted the fertile region. Should this land 

 be at length submerged, the wnvr^ of the sea may wash away 

 in a few hours the scanty covering of mould, and it may 

 merely impart a darker shade of colour to the next stratum 

 of marl, sand, or other matter newly thrown down. So also 

 at the bottom of the ocean where no sediment is accumulat- 

 ing, seaweed, zoophytes, fish, and even shells, may multiply 

 for ages and decompose, leaving no vestige of their form or 

 substance behind. Their decay, in water, although more 

 slow, is as certain and eventually as complete as in the open 

 air. Nor can they be perpetuated for indefinite periods in a 

 fossil state, unless imbedded in some matrix which is im- 

 pervious to water, or which at least does not allow a free per- 

 colation of that fluid, impregnated, as it usually is, with a 

 slight quantity of carbonic or other acid. Such a free per- 

 colation may be prevented either by the mineral nature of 

 the matrix itself, or by the superposition of an impermeable 

 stratum; but if unimpeded, the fossil shell or bone will be 

 dissolved and removed, particle after particle, and thus 

 entirely effaced, unless petrifaction or the substitution of 

 some mineral for the organic matter happen to take place. 



That there has been land as well as sea at all former 

 geological periods, we know from the fact that fossil trees 

 and terrestrial plants are imbedded in rocks of every age, 

 except those which are so ancient as to be very imperfectly 

 known to us. Occasionally lacustrine and fluviatile shells, 

 or the bones of amphibious or land reptiles, point to the same 

 conclusion. The existence of dry land at all periods of the 

 past implies, as before mentioned, the partial deposition of 

 sediment, or its limitation to certain areas; and the next 

 point to which I shall call the reader's attention is the shift- 

 ing of these areas from one region to another. 



First, then, variations in the site of sedimentary deposition 

 are brought about independently of subterranean movements. 

 There is always a slight change from year to year, or from 

 century to century. The sediment of the Rhone, for example, 

 thrown into the Lake of Geneva, is now conveyed to a spot a 

 mile and a half distant from that where it accumulated in 

 the tenth century, and six miles from the point where the 





