primitive vertebrates. The ancestral arachnids were marine forms, present in the oldest records we have; they nourished in the Cambrian, and were the highest type of animals in existence at that time. The ostracoderms flourished in the following, or Silurian, period and were the highest type of their time. They had some points in common with their predecessors, the marine arachnids, and also with the true fishes that appeared in the next, or Devonian, period, and which were likewise the highest type of their time. The inference is obvious, that the marine arachnids, the ostracoderms, and the fishes, represent three successive stages in the evolution of the animal kingdom, just as in the later periods the fishes, amphibia and mammals represent successive stages in the evolution of the vertebrates. If this inference is correct, then the whole creative period in the evolution of the vertebrate stock should become an open book, because the materials, both living and fossil, with which one can unravel the evolution of the arachnids, are apparently abundant and accessible.
This situation demanded careful investigation, for the issues at stake were very great. In 1889, when this problem for the first time assumed definite shape in my mind, it was apparently impossible to obtain well-preserved ostracoderms in this country, nor did the known remains in any country appear likely to yield more than the superficial details of their anatomy.
We were thus compelled to wait on opportunity, meantime, during the next ten or eleven years, giving our attention to the anatomy and embryology of living arachnids and the lower vertebrates, convinced that a careful study along these lines would ultimately yield definite evidence, one way or the other. The results fully justified this conclusion, for the longer this problem was studied, the more convincingly did it appear that the differences between these two great divisions of the animal kingdom were largely superficial and could be legitimately explained. The resemblances were fundamental; they were found in unexpected places, in independent systems of organs, and they ran through successive stages in the growth of those organs. It was clear that no other group of invertebrates resembled the vertebrates in such a variety of ways, or to the same extent, as did the arachnids; and no one has claimed that the main facts upon which these resemblances were based are not substantially as I have stated them to be.
We had demonstrated, therefore, that the marine arachnids are to be regarded as the most probable ancestors of the vertebrate stock.
V. The Ostracoderms
But in spite of all that, there still remained a wide gap between the arachnids and the vertebrates, and to bridge that gap we had to find
