352 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY. 
If the above arguments have validity, and I am firmly convinced 
that they do, it follows that little weight can be attached to coral 
reef faunas in attempting correlation. 
In the channels between modern reefs varied conditions obtain, 
and Dana describes the deposits as variable to a high degree, at one 
place coral sand, at another coral mud, at others clay mud with little 
coralline material and he states that “The facts show that the rocks ~ 
formed in such channels may be of all the kinds that occur in reef 
regions — coral and shell conglomerate, compact impalpable limestone, 
limestone full of Orbitolites, or containing, as well, remains of other 
species of the seas, and also rocks made of clay, mud, sand or pebbles 
of: the mountains or high lands adjoining.” ! 
As illustrative of the data given in the above quotation Dana might 
have cited the reefs of Gotland, as every word is strictly applicable. 
There, on the same level, are fine-grained limestones made from lime 
muds, limestone conglomerates, lime sandstones, clays, shell breccias, 
etc. In the passages between modern reefs the “tidal currents often 
have great strength, and are much modified and increased in certain 
places, or diminished in others, by the position of the reef with refer- 
ence to the lands” ? and there is little doubt that local erosion replaces 
deposition to a considerable extent as the growth and erosion of the 
reefs change the direction of the currents, thus developing local uncon- 
formities. In this way the many discordances of the Gotland section 
are readily explained. 
These varied conditions upon, within, and about a coral reef will 
be and are reflected in the faunas of the bottom, each species selecting 
that bottom and those conditions on and under which it best thrives, 
and animals of the same species might be found on the top of the reef 
and at its foot, as, for example, might have occurred (and where it is 
believed it did occur) at Hoburgen where about seventy-five feet sepa- 
rate the existing summit from the base. 
The sediments which are deposited on the flanks of a coral reef, 
or any similarly elevated mass, are inclined away therefrom. The 
compacting of the strata would increase the inclination while local 
slumping would intensify the undulatory structure. As a consequence 
anticlinal and synclinal structures would be developed. Some of the 
undulations of the Gotland section may, and appear to have been 
developed by subsequent movement, but it is certain that many are 
contemporaneous and are directly referable to the influence of the 
1Dana. Loc. cit., p. 152. 
2Dana. Loc. cit., p. 151. 
