commonly known as the "percentage test." The Eocene strata were defined as having only about one-thirtieth part of their Mollusca identical with living species, the Miocene as having about one-fifth, the Older Pliocene from one-third to one-half, and the Newer Pliocine nine-tenths.

The terms Eocene, Miocene, and Pliocene were at once adopted, together with the principle of classifying the Tertiary strata by means of the percentage of recent shells which they contain. But of late years several geologists and palæontologists have raised objections to the percentage test, as did Mr. Charlesworth when it was first proposed. It may not be out of place, therefore, to discuss the advantages that have accrued from its adoption during the last thirty years, and the probability of its eventually becoming superseded by some other mode of classification.

Sir Charles Lyell's classification depends on two principles; first, that the Mollusca are typical of the rest of the animal kingdom, and are, at the same time, the most convenient for the purpose; and secondly, that the percentage of recent species in a fauna varies inversely with its age. In the first place, it is certain that the fossils which are most generally and completely preserved belong to the class Mollusca. The shells of these animals are so durable, so easily recognized, and, generally speaking, so characteristic of the species to which they belong, that they form a much better medium of comparison than the remains of land-animals or of plants, which generally occur in a fragmentary condition, and the preservation of which is usually the result of some fortuitous circumstance. Again, other classes of marine organisms are either not so abundant in species and individuals, or they are not likely to be preserved in the fossil state. It is difficult to say whether the Mollusca are, or are not, typical, in their duration, of the rest of the animal kingdom. They are not so prone to change as higher organisms; but, on the other hand, species of shells do not, as a rule, exist through several geological periods, like species of Foraminifera. We should imagine, however, that while the scale furnished by each class of animals is true, each scale has a value of its own, which has a certain ratio to that of each of the rest. The scale furnished by the Mollusca being neither too large nor too small, is on this account preferable to several others; therefore, from every point of view the Mollusca seem more convenient for the purpose than any other group of

organisms.

The proposition that the greater the age of a Tertiary fauna the smaller is the proportion of recent species that it contains, is extremely difficult either to prove or to disprove; but even if it can now be shown to be untrue, it was at the time the percentage test was proposed a most convenient fiction. Sir Charles Lyell urged that the recent faunas formed a common point of departure in all

countries, and that in the event of Tertiary deposits being discovered in any region they could be referred, by means of the percentage of recent forms amongst their fossils, to their place in the Tertiary series. For many years this test has been applied with useful, even if but temporary, results; but it must also be admitted that in several cases the application has not been successful. We have Eocene and Miocene deposits in India, for instance; but the determination of the former depends almost entirely upon the fact of their containing a large number of species of Nummulites, and not upon any percentage calculation; while the reference of the latter to the Miocene division is wholly based on its Mammalian fauna. In Australia a very varied series of Tertiary deposits has been known for many years; but even the percentage test has not yet enabled Australian geologists to come to any agreement as to their Eocene, Miocene, or Pliocene date. One amateur geologist, indeed, appears to have been for years in a state of perpetual oscillation between the three.

In a series of papers Mr. Charlesworth stated thirty years ago' some of the objections which he then saw to the use of the percentage test; but although he alluded to other sources of error, he more especially dwelt on the disagreement existing between naturalists as to the amount of divergence necessary to constitute a species. To render this nugatory, he suggested an attempt to classify Tertiary strata by means of "the totality of the characters which each series exhibits," on the principle that there is a "uniform approximation to existing species, shown by the fossils of different deposits, corresponding to their respective antiquity."† But it is to be regretted that he did not himself construct the "table of degrees" which he proposed, nor illustrate his suggestion by making the attempt to classify Tertiary strata by means of it. The principle is no doubt correct, and has been used with signal success in the classification of plants into Natural Orders; it is also the one commonly used in classifying the older rocks, and ought not to be difficult of application to the Tertiary. The misfortune is that while many men possess a destructive" faculty in an eminent degree, there are so few who, like Sir Charles Lyell, are gifted with a "constructive" genius. The former class of men do not benefit science, although they show that a scheme which works well is nevertheless faulty; but the latter are entitled to our gratitude for a system which, faulty though it may be, is infinitely better than none.‡

*Mag. Nat. Hist.,' vol. ix., p. 537. Phil. Mag.,' 3rd ser., vol. vii., p. 81; vol. viii., p. 529; vol. x., p. 1.

+ Phil. Mag.,' 3rd ser., vol. x., p. 8.

"A maxim which it may be useful to recollect is this,-that hypotheses may often be of service to science, when they involve a certain portion of incompleteness, and even of error."-WHEWELL's Philosophy of the Inductive Sciences, vol. ii., p. 225.

After all, changes in physical geography and in climate, in any given region, are the chief causes of uncertainty in the application of the percentage test. The climate of Europe, for instance, was much warmer during the Eocene and Miocene periods than it is now. As the climate became colder no doubt the animals and plants which inhabited Europe migrated to warmer regions. In Europe there are many deposits of the age of these warmer periods, and it does not seem unreasonable to believe that formations in more southern latitudes, containing fossilized members of the same fauna, would be more recent in date than the apparently contemporaneous strata in Europe. Indeed, if some members of a species become modified during such a struggle for existence, as takes place in a country whose climate is becoming unsuitable for its inhabitants, while stronger individuals retain their specific characters; and if the modified form does not survive, as a species, the one from which it descended, it is easy to see that a formation containing a larger proportion of extinct species may be more recent than one containing a smaller proportion, in a different latitude, or possessing in past times a different climate.

When the percentage test was proposed, the scientific world was not ripe for the consideration of matters so calculated to disturb the principles of geological chronology, and therefore Sir Charles Lyell's scheme passed almost unchallenged. That its adoption has been attended with beneficial results is quite certain, and until some better and equally simple scheme is proposed, it will no doubt continue to be the one most generally adopted. But it behoves every philosophical geologist to remember that increase of knowledge has rendered faulty that which at one time appeared to be perfect, "inasmuch as it had the appearance of possessing arithmetical accuracy."

As science advances we are rather apt to forget that what to us are mere elementary, and apparently self-evident truths, were at one time original and great discoveries. So the services of our predecessors are not unfrequently too much underrated, and the truth of the old maxim that "familiarity breeds contempt" is proved in a new way. It seems, therefore, a good thing now and then to consider how large a debt we really do owe to those who have gone before us; often men who with imperfect aids have indicated the clue to some of nature's mysteries, which a more perfect knowledge of natural laws now enables us firmly to grasp. And if it should, as no doubt it frequently does, eventually become manifest that old ideas, interpretations, and theories are erroneous, there is not the less credit due to their authors; for have not their readings of nature for years answered all the requirements of a more perfect interpretation, and materially assisted science thereby? *Quart. Journ. Geol. Soc,' vol. xxii.,

p. 230.

For instance, the Ptolemaic and Copernican systems of Astronomy were each supplanted in its turn; but does any astronomer pretend that their authors are therefore less worthy of a place in the very front rank of the great discoverers in his science? Therefore, although we admit that the percentage test is logically not faultless, we consider that Sir Charles Lyell is not the less entitled to great credit, and exalted rank as a geologist, for its promulgation more than thirty years ago.

As we have already stated, Sir Charles Lyell gave the names. Eocene, Miocene, and Pliocene to his three great divisions of the Tertiary series; and he again subdivided the Pliocene into Older and Newer. As typical formations, he referred to the Eocene division the strata of the London and Paris Basins; to the Miocene, the Faluns of the Loire and the beds of the Superga, near Turin; to the Older Pliocene, the Crag of England and the Subapennine strata of Italy; and to the Newer Pliocene, the Sicilian beds and more recent deposits. He anticipated the future discovery of beds which would lessen the gaps that then existed between the members of these various subdivisions; and, it is almost needless to say, his anticipation has been amply realized. The question for us now to consider is how far these divisions are natural, and how far they are arbitrary. Many geologists would affirm that all divisions of strata are arbitrary, while others would contend for their being mostly natural. If the whole surface of the earth be considered, and if we are supposed to possess a complete knowledge of its geological history, then no doubt all divisions are arbitraryfor there must have been a continuous sequence of deposits. But in the present state of our knowledge-some deposits being unknown (either not explored or submerged) and others destroyed,—it is no doubt true that, for particular areas, while some divisions of strata are quite natural, others are more or less artificial. Now, into which category do these divisions of Eocene, Miocene, and Pliocene fall? Sir Charles Lyell himself would say that they are artificial, as all divisions necessarily must be. But for ourselves we should say that while these divisions are artificial, others may be, as far as Europe or any other region separately is concerned, as purely natural as any in the Geological scale.

A study of the literature of the Tertiary system will reveal the fact that in North Germany and in Austria, where certain portions of the series are extensively developed, geologists have been obliged to invent new terms to designate groups of beds which they have been unable to refer with confidence to any one of Sir Charles Lyell's divisions.

In North Germany, Professor Beyrich has grouped together, under his new term Oligocene, a long succession of beds older than the typical Miocene Faluns, and newer than the Nummulitic

(Middle Eocene) strata of the Paris Basin and other districts. Sir Charles Lyell, however, refers to his Lower Miocene all the beds as far down as the Hempstead series, including that deposit. The remaining strata in question he calls Upper Eocene. Sir Charles has confessed repeatedly that his line is purely arbitrary; but he contends that the other is equally so, and that there is consequently no need of a new term. Few Tertiary paleontologists will, we imagine, agree with him in this, and it certainly seems preferable to curtail the Eocene and Miocene, and interpolate a new group, than to be confessedly reduced to the necessity of drawing a line where there is no physical or palæontological break.

In the Vienna Basin there exists a very complete series of Miocene (Upper Miocene of Lyell) deposits, passing gradually upwards into newer strata. From the difficulty the Austrian geologists have experienced in defining the upper limit of the Miocene deposits, they have at last been led to abandon the terms Miocene and Pliocene, and to group the whole of the strata embraced within their definitions under the single term Neogene. In endeavours to assign to their place in the series the Tertiary deposits of other regions, questions have been raised as to the value of the distinction between Miocene and Pliocene strata, and some palæontologists have gone so far as to assert that the significance of the terms is far more climatal than chronological; in fact, that in tropical regions it is impossible to say that certain deposits are Miocene and not Pliocene, or vice versa. Under these circumstances it certainly does seem advisable to unite the two divisions, especially for the purpose of assigning to their proper horizon the fossils of low latitudes. The revised classification would then exhibit to the old one of Sir Charles Lyell the relation shown in the following table:

Even this revised classification cannot claim the merit of being entirely natural; but it is certainly nearer that Ultima Thule of systematists than the original one of Sir Charles Lyell. It would indeed be strange if geology had made no progress in this direction for more than thirty years; and the only marvel is that, in a science which makes such gigantic strides, the original classification has not by this time been entirely swept away. The fact that it has not, is, however, conclusive testimony of the reality and