sence of Elephas antiquus in geological deposits which are chiefly diluvial valley gravels and sand clays. These facts alone indicate a prolonged colder period, a northern or borcal climate. The fauna presents a great variety adapted to different degrees of temperature but decidedly of northern type. Other facts indicate that this colder period was initiated by a distinct second advance of the ice followed by a gradual recession, namely the occurrence of arctic and subarctic types succeeded by north temperate types, in a number of localities, typically near Schaffhausen (Schweizerbild, STEINMANN, '93, p. 117) (Franken, SCHLOSSER, '95, p. 211).

These successive northern faunas in single localities are typically as follows:

The prevailing types of this stage are the typical Elephas primigenius which succeeded Elephas trogontheri, Rhinoceros tichorhinus and Rangifer tarandus. The reindeer, first the barren ground then the woodland variety, increased rapidly in number during this period and constitute its most distinctive form; hence this is known as the Reindeer period.

It includes the most remarkable diversity of life of Asiatic both Siberian and Oriental, and of African origin. The persistence of the following southern forms: Felis (leo) spelæa: Felis pardus, Hyana (crocuta) spelæa, Equus caballus, Equus (asinus) hemiones, Rhinoceros tichorhinus (with affinities to R. simus), Elephas primigenius. All these types, excepting possibly the Mammoth, now inhabit warm, dry, semi-arid regions. There is therefore an Ethiopian and Oriental fauna, in certain localities succeeding a steppe and tundre fauna. At no period either before or since was Europe so thoroughly cosmopolitan, a fact

which has not been sufficiently emphasized previously. The climate was cold and relatively dry.

The close of this period is also the close of the Paleolithic human period which after a long interval was succeeded by the Neolithic period.

3. Upper Pleistocene, Postglacial

As above observed there is a difference of opinion as to the interglacial or postglacial age of the loess. All the North Siberian, Oriental and African types gradually disappear, the modern European forest and field fauna alone survives. There is some evidence that both the mammoth and reindeer lived for a time in this period, the latter being now confined to more northern Europe. The Irish deer, Megaceros hiberniæ, the reindeer, the bovidæ Bos taurus, Bos longifrons, Bos brachyceros, are the characteristic ruminants. Alces palmatus is a postglacial Russian moose. The horse, E. caballus, of larger and smaller varieties is now domesticated and used for food. The carnivora, rodentia and insectivora are all of modern type.

The detailed comparison of the Pleistocene of Europe, America, and Asia is still under way, and very important results may be expected from it. It will be equally serviceable to American anthropologists and palæontologists, for our own Pleistocene is far from being understood. The stages represented by our horse or Equus Beds, which are usually considered Lower Pleistocene, as well as of the Megalonyx and Cave Fauna of the East remain to be exactly fixed. Interest in this problem is greatly enhanced by the fact that we may at any moment discover the remains of man or of his ancestors associated with Equus and positively demonstrate the existence of man upon this continent at a period contemporaneous with the first proofs of his appearance in Europe in the existence of preglacial palæolithic flints.

PART II. FAUNAL RELATIONS OF EUROPE AND AMERICA DURING THE TERTIARY PERIOD AND THEORY OF THE SUCCESSIVE INVASIONS OF AN AFRICAN

FAUNA INTO EUROPE

In an address before the Academy last year the various steps which have been taken to secure correlation were described. The work proves to be a very difficult one and is by no means complete. The kind co-operation of the leading palæontologists of Europe was enlisted and as a result an approximate correlation sheet was prepared. This was virtually a report of progress in this investigation, main emphasis being laid upon geological succession. In continuing the subject this year, main emphasis will be laid upon faunal succession or the distribution of the different orders and families of mammals, concluding with the latest views as to the succession of life during the Pleistocene period in Europe.'

This portion of the second address is placed in its proper order above after

LOUP FORK

Pliocene.

Preliminary Correlation Table of European and American Tertiary Horizons. On all the levels above the Stampien the parallels are imperfectly established.

The preliminary correlation sheet abbreviated in this table sets forth the results of the geological succession and correlation so far as it has been carried at present and illustrates the rapid progress of the knowledge of our own horizons. It includes the latest results of the American Museum explorations in the Miocene of Colorado and Kansas, as roughly studied by MATTHEW, but these correlations are not to be understood as final. SCOTT has already transferred our John Day of Oregon, from the Miocene, where it was formerly placed, to the Upper Oligocene. The lower part at least of these beds belongs in the Oligocene, while the Upper John Day may prove to correspond with the Lower Miocene of Europe. Our Pliocene record as compared with the magnificent Pliocene of Europe is extremely meagre, and our Miocene succession rich as it is, is not as fully understood as the Miocene of France; we look for more exact results from the American Museum explorations which are now being collated. It is only when we pass into the great time period from the Oligocene downwards that the American record becomes a superbly complete time standard for the whole Northern Hemisphere or Holarctic Region.

II. TERTIARY GEOGRAPHICAL DISTRIBUTION

The importance of Geographical distribution was first recognized by HUMBOLDT, and set forth by DARWIN in the Origin of Species,' in 1858. In the same year SCLATER divided the world.

into six great regions and into eastern and western divisions or Palæogaa and Neogæa, to embrace the Old and New Worlds respectively, a division which has proved to be illogical. This led DARWIN's distinguished colleague, ALFRED WALLACE, to his great work upon the Geographical Distribution of Animals and the division of the world into life regions; in which Sclater's scheme was adopted and developed. In 1868 Huxley divided the world into a northern division, Arctogæa, and a southern division Notogaa to include the Northern and Southern Hemispheres respectively; this division was a little nearer the truth than Sclater's. Between 1868 and 1890, SCLATER, ALLEN, NEWTON and BLANFORD, working upon living birds and mammals, continued this investigation, but it remained for BLANFORD, in 1890, to prove that the world zoologically should be divided into three great divisions; an Australian, a South American and a third region, Arctogæa, comprising North America, Europe, Asia and Africa.2

Now it is clear that exactly as our understanding of the relations of living animals and plants to each other depends upon their fossil ancestors or upon their palæontology, so the final test of a scheme of zoological distribution must be a palæontological test. The animals of various families and orders have either originated in or migrated into their present habitat in past time, so that the geological record as to their order of appearance becomes of first importance. Here again the necessity of an absolutely reliable correlation time scale such as we are now establishing becomes evident, for the very first step toward an exact solution of the problem of past migration is to establish, as far as possible, the faunal parallels upon different continents; we can then determine where certain types of animals first appeared, and distinguish between the autocthonous endemic or native types and the migrant or new types.

The history of opinion upon this subject is fully set forth by LYDEKKER`S valuable work the "Geographical Distribution of Mammals," published in 1896.

2 Dr. Theodore Gill has kindly called attention (Science, June 8, 1900) to my oversight of an important paper of his ("On the Geographical Distribution of Fishes," Ann. Mag. Nat. Hist., 1875, pp. 251-255). He unites South America, Australia and Africa into a single division EOGÆA, in contrast with CÆNOGÆA, which includes North America, Eurasia and India.