THE LOWER LIAS OF KEYNSHAM. 
11 
the definite order A, B, C. Further, suppose that a locality is 
discovered in which A and 0 alone occur, whilst the range 
of 0 approximately starts where that of A ends. It does not 
necessarily follow that there was no deposit going on in this 
locality during the time when B was common elsewhere, but 
rather that A lived on longer and C started its existence 
earlier : the only just conclusion is that B is not valuable 
as a zonal unit outside the localities in which it actually 
occurs. 
The adoption of minute liemerce^ represented by the 
ranges of very short lived species of a single genus, must 
also be open to very considerable objection, for the fact 
that their maxima succeed each other in a definite order 
proves little more than a uniformity in the direction of evolu- 
tion ; it cannot be deduced as even probable that the actual 
period of existence of any particular species was the same 
in two distant localities, nor that the ratios of the duration 
of existence of two successive species in one locality was the 
same as for another. 
Groups, rather than species, are the best zonal indices for 
general application, whilst species are most useful for local 
comparisons. If to the name of a zonal group we add that of 
any species, we indicate a smaller sub-division of time, but, 
at the same time, we limit the geographical application, so 
that finally, when a large number of species are named as 
co-occurrent, we indicate merely the time during which a 
particular bed, to be found in a particular quarry, was laid 
down. These are the lines which we here follow ; the zonal 
indices are of world- wide application, the sub-zonal indices 
have merely local value. 
In England, the zonal series generally accepted for the 
Lower Lias, so far only as we are here concerned with it, con- 
sists of the following indices : Am. planorbis (with a sub- 
zone of Ostrea liassica), Am. angulatus, Am. hiicklandi^ 
