It seems to us that explanation must lie in the fact that mutations occur every now and again in some species. If two portions of a species are separated and a mutation occurs in one portion and not in the other, and if the mutating form succeeds in supplanting the parent form in that isolated portion of the species in which it has appeared, we should have the phenomenon of two races or species differing in appearance although subjected to what appear to be identical environment. This, of course, is pure conjecture. All that can be said of it at present is that it is not opposed to observed facts. That mutations do occur must be admitted. At present we are totally in the dark as to what causes them. They arise at the most unexpected times. In favour of the explanation based on “mutation” there is the interesting fact that geographical isolation does not by any means always cause divergence of character. This Romanes, with great fairness, freely admits. “There are,” he writes, on p. 133 of vol. iii. of _Darwin and after Darwin_, “four species of butterflies, belonging to three genera (_Lycæna donzelii_, _L. pheretes_, _Argynnis pales_, _Erebia manto_), which are identical in the polar regions and the Alps, notwithstanding that the sparse Alpine populations have been presumably separated from their parent stocks since the glacial period.” Again, there are “certain species of fresh-water crustaceans (_Apus_), the representatives of which are compelled habitually to form small isolated colonies in widely separated ponds, and nevertheless exhibit no divergence of character, although apogamy has probably lasted for centuries.” Cormorants To these examples we may add that of the cormorants. These birds have an almost worldwide range. One species—our Cormorant (_Phalacrocorax carbo_)—occurs in every imaginable kind of environment. Isolation has not effected any changes in the appearance of this species. Yet in New Zealand there exist no fewer than fourteen other species of cormorant. New Zealand is a country where climatic conditions are comparatively uniform, nevertheless it boasts of no fewer than fifteen out of the thirty-seven known species of cormorant. A possible explanation of this phenomenon may be found in the comparatively easy conditions under which cormorants live in New Zealand.[10] Under such circumstances mutants may be permitted by natural selection to survive, whereas in other parts of the world such mutants have not been able to hold their own. Prof. Bateson has likened natural selection to a competitive examination to which every organism must submit. The penalty for failure is immediate death. The standard of the examination may vary with the locality. Isolation, then, is a very important factor in the making of species, for without it, in some form, the multiplication of species is impossible.