The existence of a Silurian fauna in Ordovician times was maintained by Barrande in the case of the Bohemian basin. Lenticular patches of Silurian rocks having the lithological characters of the Silurian strata are found in the Ordovician beds of that region, and they contain fossils specifically identical with those of the Silurian rocks. Barrande explained this appearance as due to the existence of a fauna in other regions resembling the Silurian fauna of Bohemia, during the Ordovician period, when the normal Ordovician fauna of Bohemia inhabited that area. He supposed that in parts of the basin, when favourable conditions arose, colonies of the foreign fauna settled for a time, but did not get a permanent footing in the basin until the commencement of Silurian times. The theory of colonies has now been rejected for the Bohemian area, and the phenomena shown to be due to repetition of strata by folding and faulting, but it is a theory which is again and again advocated in order to explain apparently anomalous phenomena in other areas, and these apparent anomalies which are so explained, must be regarded with grave suspicion.
The various complexities alluded to in the foregoing pages increase the difficulty experienced by the geologist in correlating strata in different areas by their included organisms, but no one of them disproves the possibility of making these correlations, which can be carried on to a greater or less extent according to the nature of the faunas.
A good deal of misconception has arisen concerning the geographical distribution of former faunas, owing to the tendency to compare them exclusively with the littoral faunas of the present day. These littoral faunas have a comparatively limited geographical distribution, the forms of one marine province often differing considerably from those of an adjoining one, and still more widely from one which is remote, so that anyone confronted with the relics of faunas from the existing Australian and European seas, would find no indications furnished by identity of species that the faunas were contemporaneous. Recent researches have shown, however, that the creatures whose remains are deposited at some distance from the coast-line have a much stronger resemblance to one another than the littoral organisms have, if the fauna of two distant areas be compared. It is still a moot point which will be discussed in a later chapter, how far the deep-sea deposits of modern times are represented amongst the strata of the geological column by deposits of similar origin. But it is certain that many of the ancient strata are not littoral deposits, and it will be found that it is by comparison of the faunas of the deeper-water deposits that the geologist correlates the strata of remote regions: where shallow water deposits are formed, the faunas differ markedly in different regions, and these shallow-water forms can only be correlated owing to their occurrence between deeper-water strata. Thus if strata A, B and C be found in one area, and the fauna of A and C are deep-water forms, those of B being shallow-water forms, and in another area beds A´ contain the same fauna as A, and C´ the same fauna as C whilst the fauna of B´ is different from that of B, we can nevertheless correlate the strata B and B´ (if they be conformable with the underlying and overlying beds), because of the identity of age of the associated beds in the two areas. It will possibly be found that the strata A and C can be further subdivided into A1, A2, ... &c. C1, C2, ... by the existence of minor faunas, which are comparable in the two cases, but such subdivisions may not be established in the case of the beds B and B´.
To take actual examples:—The Llandovery beds of Dumfriesshire can be subdivided into several minor divisions each of which can be recognised in the Lake District of England, and to a large extent in Scandinavia and elsewhere, for the deposits in these areas are of deep-water character, and the sub-faunas of the subdivisions are similar in the different areas, but the Llandovery rocks of the Welsh borderland are shallow-water deposits, with a different fauna from that of the deep-water deposits of this age, and can only be stated to be contemporaneous with the Llandovery rocks elsewhere, because the deeper-water faunas of the underlying Bala rocks and overlying Wenlock rocks of the Welsh borders are respectively similar to those of the Bala and Wenlock rocks of the other regions. The shallow-water Llandoveries of the Welsh borders have only been separated into two divisions, upper and lower, and have not been split up into a number of subdivisions, each characterised by a sub-fauna, and each comparable with one of the subdivisions of Dumfriesshire, Lakeland and the other regions where the deep-water facies is found.
It will be seen that though the principle of William Smith that strata can be recognised by their included organisms has been extended since his time, and shown to apply to far smaller subdivisions of the strata than was suspected, the method of application is the same, and is more or less successful according to the amount of evidence which is accumulated in support of it.
METHODS OF CLASSIFICATION OF THE STRATA.