2 2-6 



NATURE 



17 an- 5, I 



I propose to divide the oral or glandular hypophysis 

 into two parts, viz. a duct or main oral hypophysis and a 

 glandular part or glandular hypophysis. The whole struc- 

 ture is without doubt in nearly all cases rudimentary, and 

 of little or no functional importance. A mass of informa- 

 tion bearing upon it has recently come to light in the 

 study of the developmental history of Annelids, chiefly at 

 the hands of Kleinenberg and Salensky. 



From the results of Kleinenberg's work, more especially, 

 we are placed in a position to compare the structure and 

 development of the hypophysis with those of certain 

 organs in the worms. To my mind, the comparison which 

 follows is one of the neatest in the whole range of com- 

 parative morphology ; I would therefore, before proceeding 

 further, give a brief re'swni of Kleinenberg's results so far 

 as they here concern us. 



In the first place, he records how the larval stomodasum 

 or mouth is replaced in a very complicated manner by 

 the Annelid permanent mouth or Schlimd. The latter is 

 formed as a paired involution of the stomodaeum, i.e. of 

 the epiblast, and this he considers to have originally 

 represented stomodceal glands. It encroaches upon and 

 swallows up the old mouth, and, finally fusing with the 

 hypoblast, it opens into the gut. 



The replacement of the larval mouth in Annelids by 

 a new structure was already known, but Kleinenberg 

 describes the steps of the process in great detail, and he 



-KO 



A.O. 



O.M 



V.T. 



Myxine glutimsa. Head in longitudinal section X 2. H.O, opening of 

 hypophysis; m., mouth opening; vi.t., median tooth ; K.7'. , ventral 

 teeth ; n.f., one of the folds of the nasal sensory membrane ; Bn , brain ; 

 O.M., opening of hypophysis into gut; A.O., oesophagus; N., noto- 

 chord. 



States that this mode of mouth substitution by means of a 

 paired involution is of very wide occurrence in the 

 ■Chaetopods. 



In it we have a direct parallel to the substitution 

 of the old Vertebrate mouth by a pair of gill-clefts, but, in 

 truth, we have something more. 



Another phenomenon of extreme interest is the forma- 

 tion of the special mouth (or Schlund) nervous system. 

 This apparatus is only concerned with the inttervatio/i 

 of the permanent Schlund, and takes no share in the 

 innervation of the hypoblastic alimentary canal. It 

 arises as a special process of the hinder part of the 

 suboesophageal ganglion : this grows towards the develop- 

 ing Schlund, becomes closely attached to the latter, fuses 

 with it and gets pinched off from the larger portion of the 

 suboesophageal ganglion, which is left as the first ganglion 

 of the ventral chain. 



I must here digress in order to discuss the question of 

 the presence or absence of any representative of the 

 supracESophageal ganglion of Annelids in Vertebrates,^ 

 and here again Kleinenberg comes to our assistance. 



1 have myself devoted a good deal of attention to this 

 point, and have arrived at the conclusion (held also, I 



I postpone the consideration of Prof. Semper's views on this loinf, and 

 on the nature of the mouth in Annelids and Vertebrates. 



believe, by Prof.' Dohrn) that there is no likelihood at all 

 of our finding an area in the Vertebrate brain which was 

 ever pierced by the oesophagus — pierced so as to divide 

 the brain into a supraoesophageal and a suboesophageal 

 portion, which might be compared respectively to such 

 divisions of the Annelidan nervous system. At first sight, 

 this appears like an admission that the Annelidan theory 

 of the origin of Vertebrates is untenable. But such is not 

 the case. 



From a large number of researches, including those of 

 Bergh, Salensky, and Kleinenberg, we know that the 

 supraoesophageal ganglion of Annelids certainly arises 

 independently of the ventral chain, and that it only later 

 becomes connected with the latter by the development of 

 the circumoesophageal collar. 



Kleinenberg's brilliant researches also teach us that the 

 permanent Annelidan nervous system arises through 

 substitution, and partial or entii'e disappearance of whole 

 larval nervous apparatuses and sense-organs. And, indeed, 

 after reading his beautiful work, one is fully prepared for 

 one of the closing statements in it — that possibly the 

 supraoesophageal ganglion is entirely absent in Verte- 

 brates. 



Personally, I have no hesitation at all in accepting this 

 as probably true ; but the grounds for my belief, or some 

 of them, I can only hint at here. They arise out of as yet 

 unpublished developmental researches. Briefly stated, I 

 see in the development of the gill-clefts, with their special 

 sense-organs and ganglia^ — all of which lie in the region 

 which is under the control of a system comparable to the 

 ventral nerve-cord of Annelids — a probable cause of the 

 disappearance of the supraoesophageal ganglion in the 

 ancestors of Vertebrates, — in a similar way to that in which, 

 according to Kleinenberg, the dislodging and destruction 

 of the special larval ganglionic centres takes place in the 

 Annelid. 



I believe that in the ancestors of the Vertebrates, by the 

 development of the eyes, and of the important gill sense- 

 organs and ganglia, the ventral chain came to obtain 

 control over a very extensive system of ganglia, sense- 

 organs, and muscles ; and, having already a control over 

 the mouth or Schlund, it entirely deposed the supra- 

 oesophageal ganglion (and its sense-organs). The entire 

 raison d'etre of the latter being thus disposed of, it 

 naturally degenerated and finally disappeared. 



If it be admitted that the supraoesophageal ganglion of 

 Annelids is absent in Vertebrates, and that the brain and 

 spinal cord of the latter may be compared directly with 

 the ventral cord of Annelids, then a whole host of direct 

 structural relationships between Annehda and Vertebrates 

 may be established. Kleinenberg expresses his opinion 

 that the spinal ganglia of Vertebrates have their parallel 

 in the parapodial ganglia of Annelids, — a comparison 

 which, as I shall elsewhere show, is entirely justifiable for 

 the spinal ganglia and for certain portions of the cranial 

 ganglia also. 



Let me now briefly review the conditions demanded of 

 any structures in the Vertebrate which are to be homo- 

 logized with the permanent mouth of Annelids. Such 

 ought to arise as a paired involution of epiblast (though 

 it is conceivable a priori that the paired character might 

 be lost). This involution must fuse with, and open into, 

 the cavity of the hypoblast. It must also give rise to 

 certain glands, and it must have a special nervous system 

 of its own derived fn m the hinder part of the first ventral 

 ganglion or its homologue — which nervous system must 

 supply it alone, and no other part of the alimentary canal. 

 All these co7tditions are fulfilled by the complex called 

 hypophysis cerebri. 



In at least one case (Hippocampus) the oral hypophy- 



' Ihe cranial ganglia of Vertel rates are far more complicated mcrpho 

 logic»lly than has hitherto been recognized. In additi n to parts which 

 appear to correspond morphologically to the prsterior root ganglia of the 

 spinal nerves plus the sympathetic ganglia, they also contain the special 

 ganglia which are formed in connection with the gill sense-organs. 



