"When we find abundance of examples of the young and old of many fossil species, and can trace them through their ordinary embryonic development, why should we not find examples of the links which bound the species together?"[42] But it is equally evident that each successive series ought to contain, in addition to its own characteristic or "new" species, =all the older forms which survived into any later deposits, or are now to be found living in our modern world=. Such no doubt was the idea of those of the early geological explorers who discarded Werner's onion-coat theory, and they tried to arrange their series accordingly. This reasonable demand is still recognized as good; and the principle is alluded to by Dana when he attempts to show how strata might be discovered and "proved" to be older than the present Lower Cambrian rocks.[43] It is, I say, still recognized =in theory= that the "younger" deposits ought to contain samples of the "older" types which were still surviving, in addition to their own characteristic species; but with the progress of geological discovery it has long since been found that such an arrangement was utterly impossible. Indeed, it would almost seem as if modern writers had forgotten the principle altogether. For, as already said, according to the present chronological arrangement, many kinds of invertebrates, both terrestrial and marine, occurring in comparative abundance in our modern world, are found as fossils only in the very "oldest" rocks and are =wholly absent from all the rest!!!= Others which date from "Mesozoic times" are wholly absent from the Tertiaries, though abundant in our modern world. This I regard as another crucial test of the rationality of this idea of a life succession. Of course there are certain limitations which must be borne in mind. If we find a series of beds made up largely of deep sea deposits, we cannot reasonably expect to find in them examples of all the land forms of the preceding "ages" which then survived, nor even of the shallow water types. Nor, conversely, can we demand that, in beds crowded with the remains of the great mammals and plants, and thus probably of fresh or shallow water formation, we ought to find examples of all the marine types still surviving. We now know that each level of ocean depth has its characteristic types of life, just as do the different heights on a mountain side. This doctrine of "rock facies" was, I believe, enunciated first in 1838. Edward Forbes also did much for this same idea, showing how at the present time certain faunas are confined to definite geographical limits, and particular ocean depths. Jules Marcou about 1848 applied this principle to the fossils and showed how such distinctions must have prevailed during geological time.