From these and other characteristics (especially the peculiar construction of the internal and external sexual organs in male and female) it is clear that we must conceive the whole sub-class of the Marsupials as one stem group, which has been developed from the Promammalia. From one branch of these Marsupials (possibly from more than one) the stem-forms of the higher Mammals, the Placentals, were afterwards evolved. Of the existing forms of the Marsupials, which have undergone various modifications through adaptation to different environments, the family of the opossums (Didelphida or Pedimana) seems to be the oldest and nearest to the common stem-form of the whole class. To this family belong the crab-eating opossum of Brazil (Figure 2.272) and the opossum of Virginia, on the embryology of which Selenka has given us a valuable work (cf. Figures 1.63 to 1.67 and 1.131 to 1.135). These Didelphida climb trees like the apes, grasping the branches with their hand-shaped hind feet. We may conclude from this that the stem-forms of the Primates, which we must regard as the earliest Lemurs, were evolved directly from the opossum. We must not forget, however, that the conversion of the five-toed foot into a prehensile hand is polyphyletic. By the same adaptation to climbing trees the habit of grasping their branches with the feet has in many different cases brought about that opposition of the thumb or great toe to the other toes which makes the hand prehensile. We see this in the climbing lizards (chameleon), the birds, and the tree-dwelling mammals of various orders. Some zoologists have lately advanced the opposite opinion, that the Marsupials represent a completely independent sub-class of the Mammals, with no direct relation to the Placentals, and developing independently of them from the Monotremes. But this opinion is untenable if we examine carefully the whole organisation of the three sub-classes, and do not lay the chief stress on incidental features and secondary adaptations (such as the formation of the marsupium). It is then clear that the Marsupials--viviparous Mammals without placenta--are a necessary transition from the oviparous Monotremes to the higher Placentals with chorion-villi. In this sense the Marsupial class certainly contains some of man's ancestors. CHAPTER 2.23. OUR APE ANCESTORS. The long series of animal forms which we must regard as the ancestors of our race has been confined within narrower and narrower circles as our phylogenetic inquiry has progressed. The great majority of known animals do not fall in the line of our ancestry, and even within the vertebrate stem only a small number are found to do so. In the most advanced class of the stem, the mammals, there are only a few families that belong directly to our genealogical tree. The most important of these are the apes and their predecessors, the half-apes, and the earliest Placentals (Prochoriata).