It is clear that evidence from the last-named source, when it is available, is the most important of all, since the order of succession of the various types is given by that of the rock strata in which they occur, and we can be quite certain that we are dealing, if not with the actual ancestors, at least with the forerunners of existing species. The "imperfection of the geological record," however, is so great that the organisms preserved in the fossil state represent only an insignificant part of the whole number of organisms that have lived on the globe since life began; and it is not surprising, therefore, that in many groups the study of fossils has hitherto afforded little help towards the working out of their genealogical history. Thus, among Crustacea there are many important groups such as the Copepoda, which are entirely unknown as fossils, their small and delicate bodies being ill adapted for preservation, although there is every reason to suppose that they are a very primitive and very ancient group. In many fossil Crustacea only the hard shell or carapace has been preserved, the appendages being lost or represented only by indecipherable fragments, and in some cases it is hardly possible to guess at the affinities of the animals. Further, several important groups are already represented in some of the oldest of the fossil-bearing rocks at present known, and the differentiation of these groups must have taken place in the dark ages before the record of the fossils begins. In spite of these disadvantages, however, the study of fossil Crustacea does throw considerable light on the evolution of the group, and it is likely that interesting results in this direction await future investigations. [Illustration: FIG. 81--RESTORATION OF A TRILOBITE (_Triarthrus becki_), SHOWING THE APPENDAGES. UPPER SIDE ON RIGHT, UNDER-SIDE ON LEFT. SLIGHTLY ENLARGED. (After Beecher.)]