So Agassiz says the young lepidosteus (a ganoid fish), long after it is hatched, exhibits in the form of its tail characters thus far known only among the fossil fishes of the Devonian period. The embryology of turtles throws light upon the fossil chelonians. It is already known that the embryonic changes of frogs and toads coincide with what is known of their succession in past ages. The characteristics of extinct genera of mammals exhibit everywhere indications that their living representatives in early life resemble them more than they do their own parents. A minute comparison of a young elephant with any mastodon will show this most fully, not only in the peculiarities of their teeth, but even in the proportion of their limbs, their toes, etc. It may therefore be considered as a general fact that the phases of development of all living animals correspond to the order of succession of their extinct representatives in past geological times. The above statements are quoted almost word for word from Professor Agassiz's "Essay on Classification." The larvæ of barnacles and other more degraded parasitic crustacea are almost exactly like those of Crustacea in general. The embryos of birds have a long tail containing almost or quite as many vertebræ as that of archæopteryx. But most of these never reach their full development but are absorbed into the pelvis, or into the "ploughshare" bone supporting the tail feathers. Thus older forms may be said to have retained throughout life a condition only embryonic in their higher relatives. And the natural classification gave the order not only of geological succession but also of stages of embryonic development. Thus the system of classification improved continually, although more and more intermediate forms, like archæopteryx, were discovered, and certain aberrant groups could find no permanent resting-place. But why should the generalized comprehensive forms stand at the bottom rather than the top of the systematic arrangement of their classes? Why should the system of classification coincide with the order of geologic occurrence, and this with the series of embryonic stages? Above all, why should the embryos of bird and perch form their tails by such a roundabout method? Why should the embryo of the bird have the tail of a lizard? No one could give any satisfactory explanation, although the facts were undoubted. Mr. Darwin's theory was the one impulse needed to crystallize these disconnected facts into one comprehensible whole. The connecting link was everywhere common descent, difference was due to the continual variation and divergence of their ancestors. The classification, which all were seeking, was really the ancestral tree of the animal kingdom. Forms more generalized should be placed lower down on the ancestral tree, and must have had an earlier geological occurrence because they represented more nearly the ancestors of the higher. But this explains also the facts of embryonic development.