Morphology has continually to postulate the existence of “annectant” forms, “Archi-Mollusc,” “Protosaurian,” “Protochordate,” etc.: hypothetical animals which combine the characters of those which lie near the bases of diverging lines of descent. There is nothing to guide us in the construction of these annectant forms except the progressive simplicity of structure indicated in the morphological and palæontological series. The earlier Birds had teeth, for instance, and so have the Reptiles, therefore the annectant form had teeth, and it was an animal combining the schematic morphology of both Birds and Reptiles. But just according to the value which we attach to one morphological character rather than another, so will the structure of the annectant form differ. Is, for instance, the alimentary canal of the Vertebrate the most fundamental and conservative part of its morphology: that is, is it the structure which has been most resistant to change in the course of the evolutionary process? Then we may regard the Vertebrates as having descended from some animal which was closely related to the Annelid worms. Or is the nervous system the most conservative part of the Vertebrate anatomy? If so, we may trace back the main Chordate stem to animals which included among their characters those of the most primitive Arthropods. In the one case the annectant form joins together the Vertebrate and Annelid stems, but in the other case it would join together the Vertebrate and Arthropod stems, a conclusion which a rigid application of the results of morphology would seem to make the more probable one. [Illustration: FIG. 25.] But, however this may be, we must not fail to notice that annectant forms--“Archi-Mollusc,” “Protosaurian,” “Protochordate,” and the like, are only fictions which we base on the precise importance that we attach to one part of the essential morphology of a group of animals rather than another. These hypothetical animals, and the genealogical schemes or phylogenies of which they form the roots, are conventional summaries of the results of comparative anatomy, this term being used to include the anatomy of the developing animal and that of extinct forms. So long as we do not possess a representative series of the fossil remains of the animals which have existed in the past, all schemes of descent founded on the comparison of the parts or the organs of living animals, or on the comparison of stages of development, must possess doubtful value when they profess to indicate the direction taken by evolution. Their true value lies rather in the way they epitomise our knowledge of morphology, and in the incentive which they give to sustained and minute investigation of the structure of animals. [Illustration: FIG. 26.]