October 8, 1S96] 



NA TURE 



555 



not fat tissue, and is most marked in the lowest vertebrates. 

 The natural interpretation of this phylogenetic history is that 

 the cranial cavity is too large for the brain in the lowest verte- 

 brates, and is filled up with a peculiar glandular substance 

 bec.iuse that glandular substance preexisted as a functional 

 organ or organs, and not because it was necessary to surround 

 the lirain with packing material in order to keep it steady, 

 owing to the unfortunate mistake having been made of forming 

 a brain much too small for its case. 



(2) The analoiiiiial /est shows that this glandular and pig- 

 mented material is in the same position with respect to the 

 central nervous system of .-Xmniocojles as the generative and 

 liver material with respect to the central nervous system and 

 alimentary canal of Limu'us. 



Fig. 3. — Head Rcj; 



(3) T/u- oiitogenetk test remains to Ije worked out. I do not 

 know the origin of this tissue in Ammocretes ; the evidence has 

 not yet been given by Kuppfer (" Studien z. vergleich. Entwick- 

 lungsgesch. d. Kopfes der Kranioten,'' 2 Heft, Mlinehen u. 

 Leipzig, 1S94). He has, however, shown (hat the neural ridge 

 gives origin to a mass of mesoblastic cells, the further fate of 

 which is not worked out. The whole story is very suggestive 

 from the point of view of my theory, but incomprehensible on 

 the view that the neural ridge is altogether nervous. 



Kinally, we ought to find in the invertebrate group in question 

 indications of the commencement of the enclosure of the ali- 

 mentary canal by the central nervous system ; such is, in fact, 

 the case. In the scorpion group a marked process of cephalisa- 

 tion has gi>ne on, so that the separate ganglia, both of the pro- 



NO. 1406, VOL. 54I 



somatic and mesosomatic region, have fused together, and fused 

 also with the large supra-resophageal mass. In the middle of 

 this large brain mass a small canal is seen closely surrounded 

 and compressed with nervous matter, as is shown in this speci- 

 men of Thelyphonus ; this canal is the alimentary canal. Again, 

 Hardy, in his work on the nervous system of Crustacea, has 

 sections througli the brain of Branchipus which demonstrate so 

 close an attachment between the nervous matter of the optic 

 ganglion and the anterior diverticulum of the gut that no line of 

 demarcation is visible between the cells of the gut wall and the 

 cells of the optic ganglion. 



For all these reasons I consider that the tubular nature of the 

 vertebrate central nervous system is explained by my hypothesis 

 much more satisfactorily and fully than by any other as yet put 

 forward ; it further follows that if this hypothesis enables us to 

 homologise all the otlier parts of the head region of the verte- 



Trabecul.-e 

 Pituitary Body 

 Old CEsophagus 

 Serrated Edge 

 Cili.-ited Gn 



Fig. 3.— Head Region of .\mmoccetes, split longitudinally into a ventral and 

 dorsal half. (Dorsal Half.) 



biate with similar parts in the arthropod, then it ceases to be an 

 hypothesis, but rises to the dignity of the most probable theory 

 of the origin of vertebrates. 



Origin of Segmental Cranial Nei~i'es. 

 I. The phylogenetie test. — It follows from the close resem- 

 blance of the brain region of the central nervous systems in the 

 two groups of animals that the cranial nerves of the vertebrate 

 must be homologous with the foremost nerves of such an animal 

 as Limulus, and must therefore supply homologous organs. 

 Leaving out of consideration for the present the nerves of special 

 sense, it follows that the segmental cranial nerves must be 

 divisible into two groups corresponding to two sets of segmental 

 muscles, viz. a group supplying structures homologous to the 

 appendages of Limulus and its allies, and a group supplying the 

 somatic or body muscles ; in other words we must find precisely 



