connection of geographical and geological distribution, the details of which are of the deepest interest, and have not yet been fully worked out. One great principle is, however, sufficiently established; namely, that the northern regions have been the birthplace of new forms of land life, whence they have extended themselves to the south, while the comparative isolation and equable climate of the South American and Australian regions have enabled them to shelter and retain the old moribund tribes. Those smaller portions of land separated from the continental masses, the islands properly so called, present, as might be expected, many peculiar features. Wallace divides them into two classes, though he admits that these pass into each other. Continental islands are those in the vicinity of continents. They consist of ancient as well as modern rock formations, and contain animals which indicate a former continental connection. Some of these are separated from the nearest mainland only by shallow seas or straits, and may be assumed to have become islands only in recent geological times. Others are divided from the nearest continent by very deep-water, so that they have probably been longer severed from the mainland. These contain more peculiar assemblages of animals and plants than the islands of the former class. Oceanic islands are more remote from the continents. They consist mostly of rocks belonging to the modern geological periods, and contain no animals of those classes which can migrate only by land. Such islands may be assumed never to have been connected with any continent. The study of the indigenous population of these various classes of islands affords many curious and interesting results, which Wallace has collected with vast industry and care, and which, on the whole, he explains in a judicious manner and in accordance with the facts of geology. When, however, he maintains that evolution of the Darwinian type is "the key to distribution," he departs widely from any basis of scientific fact. This becomes apparent when we consider the following results, which appear everywhere in the discussion of the various insular faunas and floras:--(1) None of these islands, however remote, can be affirmed to have been peopled by the spontaneous evolution of the higher animals or plants from lower forms. Their population is in every case not autochthonous, but derived. (2) Even in those which are most distant from the continents, and may be supposed to have been colonized in very ancient times, there is no evidence of any very important modification of their inhabitants. (3) While the facts point to the origin of most forms of terrestrial life in the Palearctic and Nearctic regions, they afford no information as to the manner or cause of their origination. In short, so far is evolution from being a key to distribution, that the whole question would become much more simple if this element were omitted altogether. A few examples may be useful