bones, teeth, and footprints preserved. Mammals, being terrestrial, are at a disadvantage. Birds, whose bodies are light enough to float, are probably devoured as food, and have less chance of preservation; they are therefore rare as fossils. These parts can, as a rule, only be preserved in a fragmentary condition. 2. Only certain deposits can so preserve them, notably mud. 3. Animals must die at such places and times that they can be preserved ; bones, &c., must be carried down regularly to a particular locality. 4. The deposit must extend over a very long time and continuously, if the series is to be complete, or even approximately so. 5. The area must be in subsidence or else it would be filled up. 6. Fossils once imbedded must be raised above the sea. 7. They must escape denudation and be exposed at some workable spot. 8. The intervals between the deposits often represent times of denudation. The denuded parts will be the uppermost-i.e., the most recent, and will break the series effectively by the removal and destruction of fossil records. All these conditions can very seldom be fulfilled. The difficulty is well illustrated by our state of knowledge with regard to domestic animals. Where are the bones of the intermediate forms between the rock-pigeon and the pouter, or fantail, for example, to be found? Yet we know that these existed but Again, to return to biology, Phenacodus (Fig. 6), one of the most important of recent fossil discoveries, was found in the Eocene of North America, and in several forms, varying in size from that of a small terrier to a leopard. This is a good example of a generalised type, having five clawed digits, a small brain, and complete radius and ulna. In many ways it suggests the ancestral form from which the Artiodactyla (deer and sheep) and the Perissodactyla (rhinoceros and horse) may have sprung; perhaps, also, it is nearly the parent form of the Carnivora. Among Birds and Reptiles a well-known series is known, due to Professor Huxley, connecting the one group with the other. Birds form a very compact and sharply limited group, characterised by their wings, feathers, and other peculiarities. In the first edition of the "Origin of Species," Darwin said: "We may thus account even for the distinctness of whole classes from each otherfor instance, of birds from all other vertebrated animals, by the belief that many animal forms of life have been utterly lost, through which the early progenitors of birds were formerly connected with the early progenitors of the other vertebrate classes." This was a prophecy out of which much capital was made at the time. It appeared an easy way out of the difficulty to suppose extinction and disappearance of all those forms whose existence, at one time or another, it was necessary to assume. At the time of the utterance of this prophecy, in 1859, there was no positive evidence at all. But in 1862, the Archæopteryx (Frontispiece) was shown to be a true bird as regards its feathers and wings, combined with several Reptilian characters, such as the long tail, the nature of the hip-bones, legs, and vertebræ. A second specimen was found in 1879, having a skull with numerous teeth, clawed fingers, perfect feathers, and bi-concave vertebræ. In 1868 Professor Huxley showed in fossil reptiles (Dinosaurus) the nature of the modification in virtue of which the quadrupedal reptile passed into the type of a bipedal bird. Again, in 1875, the discovery of toothed birds in chalk by Marsh completed the series of transitional forms between birds and reptiles. From that time Darwin's prophecy could be replaced by demonstrated facts. There are actual fossils which bridge over the E gap between reptiles and birds, in this sense that they enable us to picture to ourselves forms from which both birds and reptiles as we know them could have sprung. The Pterodactyl shows how flight is possible to a reptile, and is possibly related to birds, although this point is doubtful. (Fig. 7.) The most famous instance of geological evidence FIG. 7. Flying Reptile (Diagrammatic Figure). is found in the Horse, and, although familiar, is so important as to bear repetition. The typical number of toes and fingers is five, as in ourselves. In quadrupeds generally the number is reduced, but the horse, zebra, and ass stand alone in having only one digit on each foot, corresponding to the middle finger and toe. If we compare the foot of the horse with that of man, we find the "hock" of the horse corresponds with man's heel; the "cannon-bone" is the metacarpal; the "pasterns" form the first two phalanges, and the "coffin-bone" the terminal phalanx of the toe. The "hoof" corresponds to the nail. In the fore-limb the "knee" of the horse is equivalent to the wrist. The "splint-bones" represent the metacarpal bones of the first and third digits. Now the ancestors of the horse are Protohippus or Hipparion, which is found in the Pliocene; Miohippus and Mesohippus, found in the Miocene ; Orohippus, in the Eocene; and Eohippus, at the base of the Eocene. In Protohippus each foot has three well-formed digits; Miohippus, in addition to this, has a rudimentary metacarpal bone of a fourth digit in the fore-foot; in Mesohippus this rudimentary metacarpal bone is more fully developed; in Orohippus there are four well-developed digits in the fore-foot, three in the hind-foot; while in Eohippus five digits are present. Thus this series of fossil forms furnishes a complete gradation from the older tertiary forms with four or five toes up to the horse with one toe. These forms differ not only as regards the number of toes, but also in other respects, chiefly in the gradual diminution and loss of independence of the ulna and fibula, and in the gradual elongation of the teeth and increasing complexity of their grinding surfaces. An excellent series of gradational forms is shown in the case of Paludina, of which six or eight disconnected forms were known first, and described as distinct species; later on, connecting forms were discovered, and it was realised that we had a case of progressive modification from the older geologic beds |