In fact, in the theory of Darwin, transmutations do not take place, either by chance or in every direction. They are ruled by certain laws which are due to the organisation itself. If an organism is once modified in a given direction, it can undergo secondary or tertiary transmutations, but will still preserve the impress of the original. It is the law of _permanent characterisation_ which _alone_ permits Darwin to explain the filiation of groups, their characteristics and their numerous relations. It is by virtue of this law that _all_ the descendants of the first mollusc have been molluscs; _all_ the descendants of the first vertebrate have been vertebrates. It is clear that this constitutes one of the foundations of the doctrine. It follows that two beings belonging to two distinct types can be referred to a _common ancestor_, whose characters were not clearly developed, but the one cannot be the descendant of the other. Now man and apes present a very striking contrast _in respect to type_. Their organs, as I have already remarked, correspond almost exactly term for term; but these organs are arranged after a very different plan. In man they are so arranged that he is essentially a _walker,_ while in apes they necessitate his being a _climber_ just as strongly. There is here an anatomical and mechanical distinction which had already been proved, as regards the inferior apes, by the works of Vicq d’Azyr, Lawrence, Serres, etc. The investigations of Duvernoy on the gorilla, of Gratiolet and M. d’Alix upon the chimpanzee, have established the fact that the anthropomorphous apes possess the same fundamental character in every point. Moreover, a glance at the page where Huxley has figured side by side a human skeleton and the skeletons of the most highly developed apes, is a sufficiently convincing proof of the fact. The consequence of these facts, from the point of view of the logical application of the law of _permanent characterisation_, is that man cannot be descended from an ancestor who is already characterised as an ape, any more than a catarrhine tailless ape can be descended from a tailed catarrhine. A _walking animal cannot_ be descended from a _climbing_ one. This was clearly understood by Vogt. In placing man among the _primates_ he declares, without hesitation, that the lowest class of apes have passed the landmark (the common ancestor) from which the different types of this family have originated and diverged. We must then place the origin of man beyond the last ape if we wish to adhere to one of the laws most emphatically necessary to the Darwinian theory. We then come to the _prosimiæ_ of Haeckel, the loris, indris, etc. But these animals also are climbers; we must go further, therefore, in search of our first direct ancestor. But the genealogy traced by Haeckel brings us from the latter to the _marsupials_.