78 
THE AVONIAN OF THE AVON GORGE 
within a very limited area and must always be employed with 
great caution. 
For example the /ummosa-dolomites, the Caninia- oolite and 
the Carnmn-shales and dolomites succeed one another in ascend- 
ing order throughout the Bristol Area, but their relative thick- 
nesses vary considerably from point to point and the sequence 
cannot be recognized in the Mendip Area, which lies only a 
short distance to the south. 
On the other hand, a sequence of biozones may be of 
world-wide application and, however ill-chosen, must always 
express some portion of the truth, seeing that the definition of 
a biozone must always be the record of part of the fauna 
which existed during a definite interval of relative time. 
It is, however, easy to understand how a series of biozones, 
defined by the successive faunas of a particular locality, may 
appear to fail when applied to a more extended area. 
For, let us suppose that three successive zones at a certain 
locality indicate oscillatory succession of conditions, as expressed 
by the ascending sequence : — 
A, containing a normal shallow- water fauna ; 
B, containing a normal fauna of moderate depth ; 
C, containing a normal shallow- water fauna. 
The faunas of A and C will be essentially similar, and that of 
B markedly different, since the life-assemblage at different 
bathymetric levels is distinct. 
If, now, the fauna of C, while presenting the same broad 
generic facies, differs from that of A in the introduction of new 
species in place of old ones, it will be perfectly legitimate to 
select the faunas of A and C as defining two successive biozones. 
In so doing, we implicitly adopt shallow- water conditions as our 
standard conditions for purposes of zoning and consequently the 
complete series of biozones must express the successive shallow- 
water faunas throughout the whole interval which is to be zoned. 
The fauna of B would be very incorrectly set down as con- 
stituting a zonal assemblage intermediate between those of A 
and C, for it is impossible to discover, either how early in the 
\ 
