The VERTEBRATA.—­Among fishes I have referred to the Coelacanthini (comprising the genera Coelacanthus, Holophagus, Undina, and Macropoma) as affording an example of a persistent type; and it is most remarkable to note the smallness of the differences between any of these fishes (affecting at most the proportions of the body and fins, and the character and sculpture of the scales), notwithstanding their enormous range in time.  In all the essentials of its very peculiar structure, the Macropoma of the Chalk is identical with the Coelacanthus of the Coal.  Look at the genus Lepidotus, again, persisting without a modification of importance from the Liassic to the Eocene formations, inclusive.

Or among the Teleostei—­in what respect is the Beryx of the Chalk more embryonic, or less differentiated, than Beryx lineatus of King George’s Sound?

Or to turn to the higher Vertebrata—­in what sense are the Liassic Chelonia inferior to those which now exist?  How are the Cretaceous Ichthyosauria, Plesiosauria, or Pterosauria less embryonic, or more differentiated, species than those of the Lias?

Or lastly, in what circumstance is the Phascolotherium more embryonic, or of a more generalized type, than the modern Opossum; or a Lophiodon, or a Palaeotherium, than a modern Tapirus or Hyrax?

These examples might be almost indefinitely multiplied, but surely they are sufficient to prove that the only safe and unquestionable testimony we can procure—­positive evidence—­fails to demonstrate any sort of progressive modification towards a less embryonic, or less generalized, type in a great many groups of animals of long-continued geological existence.  In these groups there is abundant evidence of variation—­none of what is ordinarily understood as progression; and, if the known geological record is to be regarded as even any considerable fragment of the whole, it is inconceivable that any theory of a necessarily progressive development can stand, for the numerous orders and families cited afford no trace of such a process.

But it is a most remarkable fact, that, while the groups which have been mentioned, and many besides, exhibit no sign of progressive modification, there are others, coexisting with them, under the same conditions, in which more or less distinct indications of such a process seem to be traceable.  Among such indications I may remind you of the predominance of Holostome Gasteropoda in the older rocks as compared with that of Siphonostome Gasteropoda in the later.  A case less open to the objection of negative evidence, however, is that afforded by the Tetrabranchiate Cephalopoda, the forms of the shells and of the septal sutures exhibiting a certain increase of complexity in the newer genera.  Here, however, one is met at once with the occurrence of Orthoceras and Baculites at the two ends of the series, and of the fact that one of the simplest genera, Nautilus, is that which now exists.