epoch; they have talked of this deposit being contemporaneous with
that deposit, until, from our little local histories of the changes at
limited spots of the earth's surface, they have constructed a universal
history of the globe as full of wonders and portents as any other story
of antiquity.

But what does this attempt to construct a universal history of the globe
imply? It implies that we shall not only have a precise knowledge of the
events which have occurred at any particular point, but that we shall
be able to say what events, at any one spot, took place at the same time
with those at other spots.

Let us see how far that is in the nature of things practicable. Suppose
that here I make a section of the Lake of Killarney, and here the
section of another lake--that of Loch Lomond in Scotland for instance.
The rivers that flow into them are constantly carrying down deposits of
mud, and beds, or strata, are being as constantly formed, one above the
other, at the bottom of those lakes. Now, there is not a shadow of
doubt that in these two lakes the lower beds are all older than the
upper--there is no doubt about that; but what does 'this' tell us about
the age of any given bed in Loch Lomond, as compared with that of any
given bed in the Lake of Killarney? It is, indeed, obvious that if
any two sets of deposits are separated and discontinuous, there is
absolutely no means whatever given you by the nature of the deposit of
saying whether one is much younger or older than the other; but you may
say, as many have said and think, that the case is very much altered if
the beds which we are comparing are continuous. Suppose two beds of mud
hardened into rock,--A and B-are seen in section. (Fig. 5.)

[Illustration: Fig. 5.]

Well, you say, it is admitted that the lowermost bed is always the
older. Very well; B, therefore, is older than A. No doubt, 'as a whole',
it is so; or if any parts of the two beds which are in the same vertical
line are compared, it is so. But suppose you take what seems a very
natural step further, and say that the part 'a' of the bed A is younger
than the part 'b' of the bed B. Is this sound reasoning? If you find any
record of changes taking place at 'b', did they occur before any events
which took place while 'a' was being deposited? It looks all very plain
sailing, indeed, to say that they did; and yet there is no proof of
anything of the kind. As the former Director of this Institution, Sir