294 MEMOIRS OF THE NATIONAL ACADEMY OF SCIENCES. 



viz, the mouth-parts and the four pairs of legs. No author, so far as we are aware, but Patten, 

 has extended the term brain to include the post-oesophageal mass, and he apparently does this 

 in order to carry out his fancied homology of the scorpion's central nervous system with the brain 

 and spinal cord of vertebrates. 



St. Eemy has in a masterly way described the brain of the higher Arachnida (Phalangidfe, 

 Spiders, and Scorpions), restricting the term brain to two regions, the first of which he calls, after 

 Schimkewitsch, the "optic ganglion," which furnishes the oi^tic nerves only. (St. Eemy's 

 term "optic ganglion," which is not a happy one, since the same name has been used by nearly 

 all authors for the optic ganglion of insects and decapod Crustacea, viz, the mass of cells situated 

 between the optic lobes and the optic nerves, should be designated by the general name supra- 

 (Bsophageal ganglion.*) 



According to St. Remy, this region comprises "three perfectly characterized parts in all 

 the types which we have studied, the optic lobes, the posterior stratified organ, and the cerebral 

 lobes." The second region of the brain, which St. Eemyt designates the rostro-mandibular 

 ganglion, is the midbrain of Patten. This i)ortion, then, innervates the first pair of appendages 

 of Arachnida and also the rostrum, which St. Remy regards as the homologue of the labrum 

 of insects and crustaceans. 



We hold the view that the brain, or pre-cesopha.geal ganglionic mass, or prostomial nervous 

 centers of Limulus, is not the homologue of the brain of Arachnida, but only represents the first 

 region of it, the optic ganglion of Schimkewitsch and of St. Remy, and the forebrain of Pat- 

 ten. It lacks the rostro-mandibular ganglion of the arachnid brain, this ganglionic mass being 

 fused with the oesophageal ring and more intimately united with it than with the brain. Thus 

 the brain of Limulus is simpler, more primitive, and, in this respect, more like that of Peripatus 

 and Annelids than is that of Arachnida. 



We hope to show that this can be proved not only by gross dissection of the adult, but by the 

 facts and figures which we give in this paper. 



For the adult brain can be readily dissected from the oesophageal ring without cutting into 

 the brain, and as will be seen in my figures of the horizontal sections of the brain (see especially 

 PI. VI, Fig. 1, and PI. V, Figs. 20-22, g. app'), the mass of ganglionic cells supplying the nerves to the 

 first pair of appendages are behind and outside of the brain proper.l The division can also be 

 seen in the entire brain, and also in Milne-Edwards figures. Moreover, the embryonic and larval 

 conditions prove that the rostro-mandibular region in Limulus belongs with the oesophageal ring. 



Viallanes, in his admirable memoir on the brain of the locust (Annales Sc. Nat., p. 1, 1887), 

 has proposed a new classification of the parts of the brain. 



He divides the brain of insects and of <;rustacea into three segments, i. e., the protocerebrum 

 (brain of the first zoonite) ; the deutocerebrum (brain of the second zoonite), and the tritocerebrum 

 (brain of the third zoonite). 



*The term supra-cesophageal ganglion is an inexact one, as in reality it is composed of from two to four pairs 

 of ganglia, which are separate in the embryo, becoming consolidated in the adult. The term supra-oesophageal 

 gauglioiiic mass, or prostomial ganglionic mass, would be more exact, unless a briefer single word could be suggested ; 

 meanwhile the term hraiii is sufficient for ordinary use, only bearing in mind that this organ varies in the number of 

 primitive lobes in different classes of Arthropoda. 



tLa seconde region du cerveau, que nous avons d^sign^e sous le nom de ganglion rostro-mandihilaire, a une con- 

 stitution beaucoup plus simple et, par suite, beaueoup plus semblable dans tout Ic groupe. Elle Be compose d'une 

 masse nerveuso, travers^e par le tube digestif, dont la portion sus-ajsophagienne se divise en avaut en trois lobes, 

 un impair, tres petit, d'ou part sur la ligne m^diane le nerf du rostre , et deux autres plus voluniineux, disposes late- 

 raleraent, qui donnent naissance aux nerfs des chdlic6res. Ceux-(-i formeut une seule paire chez les Phalangides et 

 chez les Aranfides oil nous les avons vus se diviserbient6ten deux branches; il y en a deux paires chez les Scorpion- 

 ides, les deux nerfs de chaque e0t6 correspondant vraisemblablement aux deux branches de division du tronc iiniquo 

 des Aran^ides. Le ganglion rostro-mandibulaire donne naissance, dans ce dernier groupe, i une paire de nerfs vis- 

 c6raux. (P. 230.) 



} Different anatomists have taken different views of this subject. Dohru remarks : "Bei Limulus aber wird 

 nur das vordorsto Paar der Gliedmassen von dom oberen Sehluud-gauglion versagt; die iibrigen empfangeu ihre 

 Nerven aus der Bauchgauglienkette," p. 585. On the other baud Lankester remarks: "In the first place the brain 

 and ojsophageal collar of Scorpio are more intimately fused with one another th:in are tlie corresponding parts of 

 Limulus," and adds further on: "From the collar, then, in Scorpio as in Limulus, the nerves to all six of the 

 pediform appendages take their origin." ("Limulus an Arachnid," pp. 9 and 10). 



