The Succession of Faunas

Application of the biogenetic law to the stages so preserved should allow of the construction of a phylogeny which can then be compared with ammonites found in earlier beds. It is claimed that by this process the phylogeny of many ammonites can be made out, and that recapitulation can be observed in the general shape of the shell, the suture-line and the ornamentation, a true phylogeny requiring that the evidence from the development of these three independent characters shall be consistent. Further more it is claimed that certain types of ornament succeed one another in a definite order in all ammonites, the shell of a mem ber of a lineage in the middle of its course beginning smooth, then developing striae, then definite ribs, which become tuberculate, then spiny, and finally repeats this series of stages in the oppo site direction, or becomes smooth almost at once. In later forms ribbed and spiny stages appear earlier in the life-history until finally all of them may be skipped, the animal being smooth throughout its life. This type of evolution, which was investigated extensively by C. E. Beecher, has been supposed to occur not only in ammonites but also in gastropods, lamellibranchs and brachiopods, in all of which groups the shell preserves its early growth-stages. It has also been recognised in the development of a colony amongst coelenterates and Bryozoa.

Zaphrentis.

Out of this vast mass it is difficult to select a single case which can be regarded as conclusively established, perhaps the best is that of the development and evolution of the simple coral Zaphrentis delanouei, from the Lower Carbo niferous of Scotland. Here we have a series of forms occurring throughout a thick series of rocks which were collected on four main horizons. It is clear that the forms from the lowest horizon ascend into those at the top by gradual change, so that no definite dividing lines could be drawn within the series. If the evolu tionary series be arbitrarily divided into four stages then it is found that at the lowest horizon 69% of the specimens collected belonged to the most primitive stage, 30% to the second, and 1% to the third evolutionary stage. At the next horizon, 2,000 feet higher, less than 1% belong to the first, 3% to the second, 69% to the third, and 28% to the fourth stage. At a still higher horizon the first stage is absent altogether. The second stage is represented by less than i%, the third by 16% and the fourth and last by 83%. In the highest horizon studied this stage is attained by 95% of all the individuals. We have here an exact parallel to the evolution of a character in Micraster. But if we examine the development during life of an individual coral from one of the later horizons, we find a very complete parallel between the evolutionary series established on quite other evidence and the individual ontogeny.

It appears that the biogenetic law, though a useful tool, must be used with caution in the construction of phylogenies, and that genealogical trees made by its aid must not be used as evi dence in favour of the hypothesis itself.

Vertebrates.

The general character of the phylogenies of invertebrates which have been made by palaeontologists is re peated in the vertebrate genealogies, which rest on far more elabo rate evidence. The best case is that of the horse. (See EQUIDAE, HORSE.) The equally complete stories of the camels, dogs and titanotheres have not yet been completely published and cannot be intelligently summarised. The rhinoceroses, as shown by the work of H. F. Osborn, have a very complex history, many differ ent lineages arising in the Oligocene, and passing up into later geological periods. In all cases the individuals increase in size, and in many the descendants of hornless forms, long after their separation, independently develop horns which though varying in number and position in the different lineages are always of the characteristic rhinoceros pattern. The separate appearance of these horns is an example of a phenomenon abundantly seen amongst fossils. It implies that the mechanism present in the fertilised egg which determines what the structure of an adult animal will become, is such that it is capable of modification only in certain definite ways, and that in rhinoceroses this mechanism is of such a nature that when a horn arises it will be of a char acteristic type, quite different from the analogous structures of giraffes, deer, oxen and titanotheres.

In certain cases the same phenomenon is presented in a some what different form. The whole structure of some part of an animal may exhibit a slow and gradual change in character going on throughout the whole of its history, and this change may be unaffected by modifications of the animal's habits. One of the best examples is to be found amongst the labyrinthodont Am phibia. Here all the Carboniferous forms are round-bodied, have the roof of the mouth completely supported by bone and are aquatic. Their immediate descendants are terrestrial and the head is a little flattened, while vacuities appear between the bones in the palate. In still later forms the head and whole body are extremely flat and the palatal vacuities have become enormous. These creatures must have been entirely aquatic ; thus in them we have a persistence of direction of structural change so regular that the geological age of any specimen can be recognised with considerable accuracy, whilst the animal's habits change twice in different directions. Furthermore it can be shown that the evolu tionary structural changes pursue the same course in different families of labyrinthodonts and in two other orders of Amphibia not closely related to them.