182 



VEETEBRATA 



greater or less elaboration of the ancestors of Urochorda, and to 

 maintain even that their ancestry had reached as high a condition 

 as that shown by Craniata ; on the other hand, it does not seem 

 likely that their point of divergence from the main ancestral line 

 leading to Craniata was lower than, or even so low as, that at 

 which Amphioxus branched off. The differentiation of trunk and 

 tail by the limitation of the notochord anteriorly is a nearer ap- 

 proach to Craniate structure than that shown by Amphioxus, 

 whilst the definite development of a brain of considerable relative 

 size places Urochorda nearer to the Craniates than is Amphioxus. 

 The metameric myomeres so strongly developed in this last are 

 not absent in Urochorda, as is often maintained, but exist in a 

 rudimentary form, indicating that they had once a fuller develop- 

 ment. 



THE HEMICHOEDA. 



Char- Hemichorda comprise the single genus Balanoglossus 



acters of formerly classified by Gegenbaur as Enteropneusta, an 



Hemi- >_ 



chorda. OK /r\\ Jk , Z 



cc 



II 



FIG. 10. Balanoglossus, anatomy and development. (Modified from Bateson.) 

 A. Balanoglossus kowalewskii, Bateson; from the coast of Virginia, U.S.; 

 natural size; a, proboscis; &, collar; c, perforate region; rf, flattened diges- 

 tive region ; t , cylindrical hind region. J3. Diagram of dorsal view, showing 

 certain organs as though the body-wall were transparent. C. Diagram of a 

 vertical antero-posterior section. D. Diagram of a dorsal view to show 

 vessels and nerves by transparency. E. Diagram of a transverse section 

 through the collar. F. Larva of B. kowaleuskii diagram of horizontal sec- 

 tion. G. Vertical longitudinal section of an older larva of the same. Letter- 

 ing B-G : a, proboscis ; 6, collar ; /, nerve-tunic of proboscis ; j?, proboscis 

 pore (ciliated orifice) ; h, notochord (limited to a small tract of modified tissue 

 derived from prrcoral extension of alimentary canal) ; i, dorsal nerve-plate ; 

 k, collar-pore (right and left), opening to exterior from collar coelomjust be- 

 neath the collar ; I, continuation of dorsal nerve-plate as a nerve-cord ; ?, 

 pharyngeal perforations (gill-slits) ; ', ccelom of proboscis (anterior azygos 

 primitive coelomic pouch) ; n 2 , collar cojlom (right and left middle ccelomic 

 pouches of embryo) ; n 3 , body coelom (right and left posterior coelomic pouches 

 of embryo) ; 0, mouth ; p, ventral nerve-tunic of body-wall ; #, proboscis 

 gland ; r, strands connecting dorsal nerve-plate with outer wall of collar ; s, 

 cavity of pharynx in front of perforate region ; t, dilated part (heart) of dorsal 

 vessels within proboscis-gland; (', dorsal vessel ; , blood-vessels of body-wall 

 in section ; w, paired nerves of collar region in transverse section ; x, peri- 

 hffimal coelom, surrounding dorsal vessel in collar region; y, digestive region of 

 gut (in embryo) ; , mesoblast. H. Larva of another species of Balanoglossus, 

 known as the Tornaria larva of Johann Miiller, :md resembling an Echinoderm 

 larva, aa, pra?oral ciliated band of Toiiiaria ; &6, post-oral ditto ; cc, terminal 

 ditto; dd, mouth; ee, apical plate and sense organ;//', canal system and 

 pore; gg, gut; M,anus. 



independent phylum of the animal kingdom. They are 

 Vertebrata of worm-like form, elongate and somewhat 



flattened from above downwards. In front of the mouth 

 is a long cylindrical proboscis, and behind it a collar, the 

 free margin .of which is turned backwards, and corresponds 

 to the opercular epipleural folds of Cephalochorda and 

 Craniata. This agreement is supported by the existence 

 of a pair of collar pores opening into the coelom of the 

 collar, as the "brown funnels" of Amphioxus open into 

 the epipleural ccelom of that animal. A proboscis pore, 

 opening on the left side into the prseoral ccelom of the 

 proboscis (paired in B. kupfferi), is exactly representative 

 of the similarly placed pore which in the young Amphioxus 

 (according to Hatschek, 15) leads into the tubular organ 

 derived from the left ccelomic chamber of the praeoral 

 lobe of that animal. The whole surface of the body is 

 ciliated, as in Nemertines and Echinoderms, and as in no 

 other Vertebrates. Following the collar is a perforated 

 region of the body, gill-slits opening from the outer 

 surface into the pharynx. In the young form there is for 

 a time, as in Appendicularix and the Ascidian tadpole, 

 only one pair of gill-slits, but they subsequently increase 

 in number as the animal grows in length. They resemble 

 in form and structure those of Amphioxus. The notochord 

 (h in fig. 10) arises at the anterior end of the hypoblast 

 in the young, and grows forward, forming a support for 

 the base of the proboscis. It is limited to this very small 

 region. The cerebro-spinal nerve-cord originates by a 

 delamination of a solid cord of epiblast in the mid- 

 dorsal line of the middle third of the body; then by 

 invagination of its two ends it extends as a tube both 

 anteriorly and posteriorly. A general network of nerve- 

 fibres (and cells?) exists beneath the epidermis all over 

 the body. The blood-system is peculiar, consisting of an 

 anterior heart and a dorsal and ventral vessel ; these are 

 united by a plexus of subcutaneous vessels. The muscu- 

 lature of the body-wall is not broken into succsssive 

 myomeres ; but, on the other hand, the gonads (ovaries or 

 testes) are sac-like, and, as in Amphioxus, are repeated in 

 a series throughout a great length of the body. In the 

 pharyngeal region the gonad sacs agree in number with 

 the gill-slits. There are no nephridia (unless proboscis 

 pore and collar pores are to be so regarded) ; but the con- 

 nective-tissue cells of the body-cavity are active as excret- 

 ing agents, as in Echinoderms and in Urochorda, and a 

 large glandular organ in the proboscis attached to the end 

 of the notochord appears to have to do with this function. 

 Not the least remarkable fact about Hemichorda is the 

 nature of their larvae. No other Vertebrata present larval 

 forms which indicate the nature of the early ancestral 

 history in what we may call pras-chordal times ; however 

 interesting the Ascidian larva, or the young Amphioxus, 

 and the embryo dog-fish, they do not take us out of the 

 Vertebrate area. Some species of Balanoglossus (? B. minu- 

 tus), however, pass through a banded ciliate larval con- 

 dition, which was known as Tornaria, and was considered 

 to be an Echinoderm larva allied to Biplnnaria, before 

 its relation to Balanoglossus was discovered. It is not 

 possible to view the Tornaria larva of Balanoglossus as 

 otherwise than identical with Echinoderm larvce, and it 

 results that Balanoglossus and the Echinoderms have 

 remote genetic affinities of a special kind. 



No classification of Hemichorda is possible beyond an Species 

 enumeration of the species : of Hem 



1. Balanoglossus clavigerus (Delia Chiaje), Naples. 



2. B. minutus (Kowalewsky), 



3. B. kowalcwskii (Al. Agassiz), cast coast, United States. 



4. B. brooksii (Bateson), ,, ,, 



5. B, salmoneus (Giard), Brittany. 



6. P. robinii (Giard), ,, 



It seems that in Balanoylossus we at last find a form 

 which, though no doubt specialized for its burrowing 



chorda. 



