186 VERTEBRATA Relation- ancestry. Its mode of life (burrowing in the sand in shallow water, ship of whilst its general build is that of a swimming animal) and the Cephalo- nature of its food (diatoms, &c., carried into the pharynx by ciliary chorda to currents) in themselves suggest such a history. The vascular Craniata system is elaborate in plan yet incomplete in detail, suggesting an and the atrophy of its finer branches, which is consistent with the small Verte- size of Amphioxus and the general principle that a complex vascular brate system can only be developed in an animal which has attained to ancestry, a certain bulk. The absence of well-developed sense organs and of " cephalization " in an animal which has attained to such elabora tion of structure as is shown by the pharynx and atrial chamber, and which has such well-developed muscles to the body-wall, is an inconsistency best explicable by degeneration ; so, too, the existence of the elaborats series of fin rays, which are out of proportion to the mechanical requirements of so small a form. Degenerate though Amphioxus must be, the ancestor from which it started on its retrogressive course was probably a long way behind any living Craniate. There is no reason to suppose that this ancestor had a cranium, or that the muscular segments and segmental nerves in its cephalic region were fused and welded. Antyhioxus has probably lost, as compared with that ancestor, lateral eyes and otocysts, nephridia, and, above all, size. The epipleural folds which now form oral hood, branchial opercula, and coalesced ventral fin were probably originally less developed lateral ridges, protecting the gill-slits anteriorly and posteriorly, serving by their undulations to assist in locomotion, whilst the median fin and its rays were large and functional. One of the most curious features in the structure of Amphioxus is its asymmetry. The anus is on the animal s right side, the nasal pit upon its left ; the myomeres on the two sides of the noto- chord do not coincide ; and the right and left dorsal spinal nerves do not arise vis-a-vis to one another. There is no conclusive reason for regarding this as an ancestral feature, although the early larval form is as curiously asymmetrical as the adult. Amphioxus habitu ally rests upon the sand, lying upon one side of the body, and it is possible that the distortion is related to this habit, as in the case of the Pleuronectid Fishes. However we may estimate Amphioxus, we are not led by it, though its muscular metamerism is so well marked, a single step in the direction of the Annelids, neither are we led directly, it is true, in the direction of Nemertina in connexion with those points, as to relationship of notochord with proboscis sheath and nerve-cord with median dorsal nerve, insisted on by Hubrecht. But it will be seen below that, by the agreement of Amphioxus with Balanoglossus in the structure of the perforations of the pharynx, in the possession of collar pores, and in the prseoral glandular body, we do arrive at an important connexion with Nemertine-like forms. THE UEOCHORDA. Char- Urochorda are Vertcbrata which, with the exception of the group acters Larvalia (Appendicularia, Fritillaria, Oikopleura), have receded of Uro- very far indeed from the characteristic Vertebrate structure, show- chorda. ing neither notochord nor nerve-cord, and gill-slits only of the most highly modified and aberrant form ; some, however (certain Ascidians), pass through a larval condition in which these struc tures are present in the normal form. It is necessary for the pur poses of the present article to confine our attention to Larvalia and to the larval forms which retain ancestral characters. (For a de scription of the whole group, see the article TUNICATA. ) In Uro chorda thus signalized the notochord never reaches forward into the anterior part of the body, but is confined to the tail (hence Uro chorda). The longitudinal muscles of the region traversed by the notochord show traces of metameric segmentation, which are prob ably survivals of a more complete development of inyorneres in ancestral forms (16). There is no trace in Larvalia of fin rays or other skeletal structure. Corresponding to the opercular folds and epipleura of Craniata and Cephalochorda are ridges of the body-wall, which protect the pharyngeal gill-slits, and may give rise, as in Cephalochorda, to an enclosed atrial chamber with atriopore. The gill-slits in these larval forms are few in number (one or two pairs), but in many of the aberrant Urochorda (by far the majority of the group) they become excessively numerous and complicated in structure, and are supported by a chitinous (?) framework, as in Cephalochorda. It has been suggested that the fenestrated structure of the pharyngeal wall in Tunicata does not represent a series of gill-slits, but a single pair of slits subdivided. This suggestion is worthy of further consideration. The cerebro-spinal nerve-cord is tubular and presents itself as a dilated cerebral vesicle in front of the notochord, and as a narrower part running along the whole length of the notochord. Sense organs are present a single eye with pigment and lens, a single otocyst, and an olfactory pit (Larvalia). The mouth is dorsal in position in the Ascidian tadpole, but subterrninal in Lar valia. The pharynx is wide, and is followed by a narrow oeso phagus, stomach, and intestine, which does not open ventrally but turns upwards to the anus. The Larvalia have a rudimentary heart and no vascular system, a fact connected with their diminu tive size. For the same reason no vascular system develops in the Ascidian tadpole until it has ceased to be locomotive and has entered upon its later development ; but in the larger adult Urochorda a contractile heart and a well-developed vascular system are present. No undeniable nephridia are present in Larvalia nor in the larval Ascidian, and no structure comparable to the collar pores of Balano glossus or the atrio-ccelomic funnels of Amphioxus is known in them. The subneural gland, however, a glandular tube opening anteriorly near the mouth of the pharynx, appears to be identical with the praeoral larval gland of Amphioxus and the proboscis pore and gland of Balanoglossus. It is probably to be regarded as a nephridium, and has been compared by Julin and Van Beneden to the pituitary body of Craniata, with which it corresponds in position and de velopment. The gonads of Larvalia are developed in irregular masses on the walls of the ccelom, ovaries and testes in the same individual. As above indicated, there is a small section of Urochorda which Classi- retain in adult life the tadpole-like form and the essential Verte- fieation brate organs which are exhibited by the larvae only of other Uro- of Uro chorda, and by a few only of these. This necessitates a primary chwda. division of the branch into two grades. Grade A. LARVALIA (Appendicularia, Fritillaria, Oikopleura). Grade B. SACCATA. Class I. Ascidise, (Simpliccs, Sociales, Composite, Pyroso- miidea). Class II. Salpiformia (Salpiidca, Dolioliidca). PIG. 9. Fritillaria (Appendicularia) furcata, one of the Urochorda. (Original drawings.) A. Lateral surface view, showing habitual carriage of "body" at right angles to " tail." B. Organs of body as seen by transparency. C. Lateral view of body with tail in morphological position, showing organs by transparency. I). Surface view of animal from below to show apertures, a, Otocyst in connexion with brain ; 6, olfactory pit; c, dorsal hood ; d, nerve- tube passing from enlarged brain to caudal region, where it forms one true ganglion and a series of minor enlargements, corresponding to the rudimentary " rnyotomes " or " myomeres " of the tail ; e, stomach ; /, ovary ; g, testis ; Ji, notochord (urochord) ; i, nerve-tube or myelon in tail ; Tc, fifth myomere of tail ; I, anus ; m, heart ; n, gill-slit ; o, endostyle or hypobranchial groove ; p, mouth. Urochorda are so extremely aberrant, and show so little more Relations than a transient developmental indication of the essential Verte- of Uro- brate organs, that we cannot hope to get much positive information chordaio from them on the subject of Vertebrate ancestry. Only the minute Verte- Appcndicularise. (Larvalia) retain the Vertebrate structure through brate life, and they are obviously, on account of their minute size, ex- ancestry,
tremely degenerate. It is possible to make hypotheses as to the