equivalent and that the division is not logical. Cuvier showed that there were four groups in the animal kingdom, each corresponding to a definite type or plan of structure, and that craniate vertebrates composed only one of these groups, invertebrates including three. In the progress of zoology it has become clear that the coelomate animals fall into a very large number of distinct groups or types, and that the vertebrates are only one class amongst many morphologically distinct classes. It has been shown further that amongst the animals that Lamarck would have placed in the Invertebrata there are several which, although devoid of vertebrae or cranium, must be associated with vertebrates in any natural system. Closer investigation of the anatomy and embryology of the craniate vertebrates showed that the possession of a jointed vertebral column was not a fundamental characteristic of the group. In some creatures, such as sturgeons and lampreys, the position of the jointed vertebral column is occupied by an unjointed rod, the so-called notochord, whilst all the Vertebrata pass through an embryonic stage in which a similar elastic unjointed notochord exists as the precursor of the jointed column. It was further found that all the vertebrates of Lamarck displayed either in the embryonic condition alone, or both in embryonic and adult conditions, a set of passages leading from the anterior-lateral portion of the body into the cavity of the pharynx, and known as gill-slits, because in those creatures in which they become functional for aquatic respiration they lodge the gills or branchial tufts. Further, it was found that in all vertebrates the great central mass of the nervous system, known as the brain and spinal cord, is in reality a hollow tube with more or less thickened walls, developed as a strand of tissue along the dorsal surface of the embryo, which sinks downwards and inwards to form a hollow tube lying dorsal to the notochord.
In 1866 A. Kowalewsky, in a memoir that is one of the classics of vertebrate morphology, worked out the development of Amphioxus, then recognized as the simplest of the vertebrate group, and compared it with the development of an Ascidian, one of a group then termed Tunicate Mollusca, and showed that the latter creature, in its larval stage, possessed, like Amphioxus, a notochord, gill-slits and a hollow dorsally placed nerve-tube. In 1877 E. Ray Lankester published a classification of the animal kingdom in which he definitely associated all the Tunicates with the vertebrates, and subdivided Vertebrata as follows: Branch A., Urochorda, which contained the Tunicates and was characterized by the limitation of the notochord to the caudal region; Branch B., Cephalochorda, containing Amphioxus, in which the notochord extended from the extreme tip of the tail to that of the snout; Branch C., Craniata, containing the Cyclostomes, Pisces, Batrachia, Reptilia, Aves and Mammalia, in which the anterior extremity of the notochord ended in the base of a cranium. Later, F. M. Balfour adopted the system of Lankester, but proposed to replace the term Vertebrata, which was anatomically misleading, by the new term Chordata, as the latter term laid stress on the existence of the notochord as the fundamental character of the group. A. Kowalewsky had shown as early as 1866 that the marine worm Balanoglossus, described by Della Chiaje at the end of the 18th century, possessed a set of gill-slits similar to those of Amphioxus and Tunicates. From 1884 to 1886 W. Bateson published a series of studies in which he suggested that there was present in Balanoglossus a representative of the notochord, and that a portion at least of its nervous system was a hollow, dorsally placed tube. On these grounds, coupled with the presence of gill-slits, he proposed to add yet a lower branch to the Chordata, to include Balanoglossus and to be termed Hemichorda, but neither Bateson nor zoologists who have written since have accepted the vertebrate affinities of Balanoglossus with complete confidence. Still more diffidently, S. F. Harmer and others have suggested that Cephalodiscus and Phoronis, still more lowly marine invertebrates, have claims to be associated with the Chordata.
It may be accepted definitely that Amphioxus and the Tunicates must be associated with the craniate vertebrates of Lamarck. With regard to the terms Vertebrata and Chordata, usage still differs. Those who wish to make the names of the larger groups significant labels prefer the term Chordata, and on the whole seem to be prevailing, but there remain many zoologists who prefer the designation with historical associations, and regard it as immaterial if, in the advance of knowledge, the connotation may have been so changed that the term has become conventional rather than verbally significant.
The characters and affinities of the lower groups that have been included under Chordata are discussed in the articles Hemichorda, Balanoglossus, Phoronidea, Pterobranchia, Tunicata and Amphioxus, so that it is necessary here to deal only with the general characters of the Chordata or Vertebrata Craniata, and to consider the views that have been advanced with regard to the origin of vertebrates.
The Vertebrata Craniata share with the Cephalochordata the fundamental characters of the group Chordata. They are bilaterally symmetrical animals with a well-marked metameric segmentation of the muscles and muscle septa, with a gut opening by an anterior ventral mouth, with lateral gill-slits in the embryo or adult, and with a ventro-posterior anus; with a dorsal tubular central nervous system, under which lies in the embryo or adult an unsegmented notochord of endodermal origin; with the body prolonged posteriorly to the anus to form a metamerically segmented tail containing notochord, nervous system and muscles; with a spacious coelomic cavity and separate blood-vascular system. They differ from the Cephalochordata in the extreme cephalization of the anterior segments of the body, including the formation of an enlarged brain with paired sense organs, the nose, eyes and auditory apparatus, and the formation of a cranium, and in the structure of the skeleton, heart, liver and organs of excretion and reproduction.
Evidence points to the origin of the Cephalochordata and the Craniata from a common ancestor in which metameric segmentation of the mesoblast and the nervous system was complete and regular. This condition has been retained by Amphioxus, but in the Craniata has been much modified. The lateral mesoblastic plates with their contained coelom are unsegmented in craniates, although traces of the primitive segmentation are visible in the development of Cyclostomes. The dorsal mesoblastic somites with the segmental musculature derived from them retain the segmental condition in Amphioxus and in the trunk region of craniates, but in the head region of the latter there has taken place a fusion or cephalization more pronounced in the higher forms, where the head is distinct from the trunk, than in lower forms where the head passes gradually into the trunk. The exact number of somites which have been cephalized is difficult to estimate, and certainly varies in different cases, but it appears to be certain that three, immediately anterior to the otic region, have been transformed into the optic muscles. Those behind the otic region (metaotic somites) vary from nine to eleven, and in Cyclostomes give rise to segmental muscles in series with those of the trunk. In true fish and higher Craniates the anterior one or two of these metaotic somites practically disappear, whilst of the remainder none form complete segmental muscles, but various portions of them give rise to muscles associated with the branchial apparatus (epibranchial and hypobranchial), the dorsal portions fading away. In other words, the metameric series continued from the trunk to the anterior end of the body in the ancestral form, retained by the Cephalochorda, and of which traces remain in the development of the Craniata, has been modified in the adult Craniata by the suppression of certain portions and the specialization of other portions to form an unsegmented structure. The process of cephalization, with, however, less complete destruction of the segmental arrangement, has also affected the anterior nerves of Craniata and brought about the distinction between cranial and spinal nerves which is a feature of the Craniates. The ancestral form must be supposed to have given off from its central nervous system lateral nerves segmentally arranged in pairs. Each member of each pair possessed two roots, a dorsal and a ventral root, possibly remaining separate, as in the Cephalochordata and the cranial nerves of Craniata, possibly joining to form a common trunk, as in the spinal nerves of Craniata. The ventral roots consisted of motor fibres passing straight outwards to innervate the segmental muscles derived from the dorsal somites; the dorsal roots took a longer course, arching outwards and round the body to supply the visceral muscles, the mucous membranes, the skin and the sense organs connected with these. It appears, moreover, that the ventral roots remained in strict association with the muscular somites to which they corresponded, and wandered beyond their own segmental areas only with these muscles, whereas the ramifications of the dorsal fibres had a wider range and were less closely bound to segmental regions. Such a primitive condition has been retained by Amphioxus, but in the case of Craniata only by the spinal nerves. Almost every great anatomist has contributed to working out the history of the cranial nerves, and it would be a hopeless task to make a just allocation of credit for the various
