is allied to the Menopoma living in North America. Species of frog (Rana), and Palæophrius an extinct genus of toads, have been found in the Miocene deposits of Germany and Switzerland.

Fresh water fish are almost unknown in the Tertiary deposits of Europe, although most of the families and some genera of living marine fish are represented from the Eocene downwards.

ANTIQUITY OF THE GENERA OF INSECTS.

Fossil insects are far too rarely found, to aid us in our determination of difficult questions of geographical distribution; but in discussing these questions it will be important to know, whether we are to look upon the existing generic forms of insects as of great or small antiquity, compared with the higher vertebrates; and to decide this question the materials at our command are ample.

The conditions requisite for the preservation of insects in a fossil state are no doubt very local and peculiar; the result being, that it is only at long intervals in the geological record that we meet with remains of insects in a recognisable condition. None appear to have been found in the Pliocene formation; but in the Upper Miocene of Eninghen in Switzerland, associated with the wonderfully rich fossil flora, are found immense quantities of insects. Prof. Heer examined more than 5,000 specimens belonging to over 800 species, and many have been found in other localities in Switzerland; so that more than 1,300 species of Miocene insects have now been determined. Most of the orders are represented, but the beetles (Coleoptera) are far the most abundant. Almost all belong to existing genera, and the majority of these genera now inhabit Europe, only three or four being exclusively Indian, African, or American.

In the Lower Miocene of Croatia there is another rich deposit of insects, somewhat more tropical in character, comprising large white-ants and dragon-flies differently marked from any

now inhabiting Europe. A butterfly is also well preserved, with all the markings of the wings; and it seems to be a Junonia, a tropical genus, though it may be a Vanessa, which is European, but the fossil most resembles Indian species of Junonia.

The Eocene formations seem to have produced no insect remains; but they occur again in the Upper Cretaceous at Aix-la-Chapelle, where two butterflies have been found, Cyllo sepulta and Satyrites Reynesii, both belonging to the Satyridæ, and the former to a genus now spread over Africa, India, and Australia.

A little earlier, in the Wealden formation of our own country, numerous insects have been found, principally dragon flies (Libellula, Æshna); aquatic Hemiptera (Velia, Hydrometra); crickets cockroaches, and cicadas, of familiar types.

Further back in the Upper Oolite of Bavaria-which produced the wonderful long-tailed bird, Archaeopteryx-insects of all orders have been found, including a moth referred to the existing genus Sphinx.

In the Lower Oolite of Oxfordshire many fossil beetles have been found whose affinities are shown by their names:-Buprestidium, Curculionidium, Blapsidium, Melolonthidium, and Prionidium; a wing of a butterfly has also been found, allied to the Brassolidæ now confined to tropical America, and named Palaontina oolitica.

Still more remote are the insects of the Lias of Gloucestershire, yet they too can be referred to well-known family typesCarabidæ, Melolonthida, Telephoridæ, Elateridæ, and Curculionidæ, ainong beetles; Gryllida and Blattidæ among Orthoptera; with Libellula, Agrion, Eshna, Ephemera, and some extinct genera. When we consider that almost the only vertebrata of this period were huge Saurian repliles like the Icthyosaurus, Plesiosaurus, and Dinosaurus, with the flying Pterodactyles; and that the great mass of our existing genera, and even families, of fish and reptiles had almost certainly not come into existence, we see at once that types of insect-form are, proportionately, far more ancient. At this remote epoch we find the chief family types (the genera of the time of Linnæus) perfectly differentiated

and recognisable. It is only when we go further back still, into the Paleozoic formations, that the insect forms begin to show that generalization of type which renders it impossible to classify them in any existing groups. Yet even in the coal formation of Nova Scotia and Durham, the fossil insects are said by competent entomologists to be "allied to Ephemera," "near Blatta," "near Phasmide;" and in deposits of the same age at Saarbrück near Trèves, a well-preserved wing of a grasshopper or locust has been found, as well as a beetle referred to the Scarabeidæ. More remarkable, however, is the recent discovery in the carboniferous shales of Belgium, of the clearly-defined wing of a large moth (Breyeria borinensis), closely resembling some of the Saturniidæ ; so that we have now all the chief orders of Insects-including those supposed to be the most highly developed and the most recent well represented at this very remote epoch. Even the oldest insects, from the Devonian rocks of North America, can mostly be classed as Neuroptera or Myriapoda, but appear to form new families.

We may consider it, therefore, as proved, that many of the larger and more important genera of insects date back to the beginning of the Tertiary period, or perhaps beyond it; but the family types are far older, and must have been differentiated very early in the Secondary period, while some of them perhaps go back to Palæozoic times. The great comparative antiquity of the genera is however the important fact for us, and we shall have occasion often to refer to it, in endeavouring to ascertain the true bearing of the facts of insect distribution, as elucidating or invalidating the conclusions arrived at from a study of the distribution of the higher animals.

ANTIQUITY OF THE GENERA OF LAND AND FRESH-WATER

SHELLS.

The remains of land and fresh-water shells are not much more frequent than those of insects. Like them, too, their forms are very stable, continuing unchanged through several geological

periods. In the Pliocene and Miocene formations, most of the shells are very similar to living species, and some are quite identical. In the Eocene we meet with ordinary forms of the genera Helix, Clausilia, Pupa, Bulimus, Glandina, Cyclostoma, Megalostoma, Planorbis, Paludina, and Limnæa, some resembling European species, others more like tropical forms. A British Eocene species of Helix is still living in Texas; and in the South of France are found species of the Brazilian sub-genera Megaspira and Anastoma. In the secondary formation no true land shells have been found, but fresh water shells are tolerably abundant, and almost all are still of living forms. In the Wealden (Lower Cretaceous) and Purbeck (Upper Oolite) are found Unio, Melania, Paludina, Planorbis, and Limnoa; while the last named genus occurs even in the Lias.

The notion that land shells were really not in existence during the secondary period is, however, proved to be erroneous by the startling discovery, in the Palæozoic coal measures of Nova Scotia, of two species of Helicidæ, both of living genera-Pupa vetusta, and Zonites priscus. They have been found in the hollow trunk of a Sigillaria, and in great quantities in a bed full of Stigmarian rootlets. The most minute examination detects no important differences of form or of microscopic structure, between these shells and living species of the same genera! These mollusca were the contemporaries of Labyrinthodonts and strange Ganoid fishes, which formed almost the whole vertebrate fauna. This unexpected discovery renders it almost certain, that numbers of other existing genera, of which we have found no traces, lived with these two through the whole secondary period; and we are thus obliged to assume as a probability, that any particular genus has lived through a long succession of geological ages. In estimating the importance of any peculiarities or anomalies in the geographical distribution of land shells as compared with the higher vertebrates, we shall, therefore, have to keep this possible, and even probable high antiquity, constantly in mind.

We have now concluded our sketch of Tertiary Palæontology as a preparation for the intelligent study of the Geographical VOL. I.-13

Distribution of Land Animals; and however imperfectly the task has been performed, the reader will at all events have been convinced that some such preliminary investigation is an essential and most important part of our work. So much of paleontology is at present tentative and conjectural, that in combining the information derived from numerous writers, many errors of detail must have been made. The main conclusions have, however, been drawn from as large a basis of facts as possible; and although fresh discoveries may show that our views as to the past history of some of the less important genera or families are erroneous, they can hardly invalidate our results to any important degree, either as regards the intercommunications between separate regions in the various geological epochs, or as to the centres from which some of the more important groups have been dispersed.