304 MEMOIKS OF THE NATIONAL ACADEMY OF SCIENCES. 



*. e., two pairs of eye-lobes or gancclia, viz. — first, those wliich send nerves to tlie lateral eyes; see- 

 oiid, those which innervate the median eyes; and, third, the cerebral ganglia, which, with the 

 mushroom bodies, form the chief portion of the brain. We have seen that the ganglia of the first 

 pair of appendages are separate from the prestomial or prinesophageal brain mass. Hence we 

 throwout these ganglia in our consideration of the constituents of the brain proper. 



Prof. Patton, though he has not published any observations on the structure of the brain of the 

 adult Limulus, in liis article on the "Origin of Vertebrates from Arachnids" briefly describes (pp. 

 336-3^6) the development of the "cephalic lobes" and of the eyes of the embryo of Limulus. 

 Patton does not give details of the structure of the brain of the embryo, only stating such facts as 

 have a bearing on iiis remarkable theoretical conclusions,* reserving a full description for a future 

 paper. On pages 338 and 339, however, he gives " diagrams " and a " semidiagrammatic" view of the 

 brain of the embryo Limulus and refers to two pairs of median eyes, with their neuromeres, and 

 on page 343 gives a figure of the "fore and mid brain of young larva of Limulus (just hatched) 

 seen fi-om neural surface," in which four prewsophageal neuromeres with their nerves are repre- 

 sented. Deferring to a few pages further on in this essay, the consideration of the number of 

 brain neuromeres in the embryo and larva of Limulus, we will simply remark that we have, after 

 repeated examinations, been unable to find more than a single pair of median eye^st in the embryo 

 or larval Limulus, nor have we been able to detect more than three pairs of neuromeres in front 

 of those innervating the first pair of appendages, these three pairs of neuromeres finally consti- 

 tuting the brain. 



Our studies as regards the median-eye lobes entirely confirm those published by Kishinyoue, 

 in his excellent work on the development of the Japanese Limulus longispina. He remarks: 



About two weeks before the hatching of the embryo the brain proper becomes divided into two transverse por- 

 tions by a constriction (Fig. 69). Thus the brain of Limulus may be divided into four parts— the ganglion of the 

 median eyes, the ganglion of the lateral eyes, and the anterior and posterior portions of the brain proper. 



These last two portions form apparently the cerebral lobes and mushroom bodies. Mr. 

 Kishinyoue has also written me, since his paper was published, that he had only observed a single 

 liair of median eyes. 



The only other author besides myself who has studied the brain of the adult Limulus is M. 

 Viallanes. An abstract of his paper read before the French Academy, December 1, 1890, appeared 

 in the Journal of the Royal Microscopical Society, fi)r February, 1891. 



The author considers that the brain in Limulus pohjphcmus (this being the species investi- 

 gated by him) " gives origin to the ocellar nerve," to the nerve of the compound eye, to the cheli- 

 ceral nerve, to the stomatogastric nerve." We copy his description of the brain : 



structure du cerveau. — Le cerveau se compose de deux paires de centres ganglionnaires. La premiere est essen- 

 tiellement constitute comnie \e protocerebron des autres arthropodes et doit ctre d<isign(;e sous ce nora. Quant h, la 

 seconde, quelques doutes pouvant encore subsister sur ses homologies, je la designerai provisoirement sous le nom de 

 cerveau posterieur. 



Protoc^rehroH.— 11 se compose d'une paire de nodules fibrenx on lobes protoc6rebraux relativement petite, ri^unis 

 sur la ligno m^diane par une commissure priaaophagieune. Les lobes protociSr6braux sont partiellemeut rovetus 

 d'une ^corce de graudes cellules uuipolaires ; et chacun d'eux donne naissanc eau nerf ' de I'ocelle correspondant. 



Quant au nerf de I'ttul coniposi?, il s'unit au lobe protoc6r(5bral correspondant, non pas directement, mais par 

 I'intermediaire d'une formation comparable dans ses traits essentiels au lobe optique des inseetes et des crustac<5s, 

 c:ir on y reconnalt les homologues de la lame gauglionuaire, du chiasma externe, des masses medullaires externe 

 et interne ; niais chez le Limule co lobe optique est trcs petit relativement et enfoui au sein de la masse du cerveau, 

 au lieu d'dtre ^cart6 de cclui-ci et en contact imm(5diat avec I'a-il composd. Chez le Limule, le nerf optique est done 

 constitu^ par des fibres postr<5tinienues ^tiroes a Textreme. 



A chacun des lobes protocdrdbraux est annexe un organe qui, en raison de ses rapports anatomiques et de sa 

 structure histologique, doit etre assimil6 au corps pedonculii des inseetes. 



•With Patton's attempt to derive the vertebrates from forms comparatively high, so specialized and modified 

 as the scorpions and spiders, to say nothing of Limulus, we have little symp.athy, regarding them not only as 

 unsound, but as tending to lead to inaccurate observations. On general grounds the attempt to derive the \-ertebrates 

 from any but the most primitive and generalized vermian forms would seem to be hazardous, the Arthropoda being 

 a specialized and completed branch of the animal series. 



t Patton considers that a diverticulum of the median eycbnlb "represents, in all probability, a pair of eyes 

 belonging to the first brain segment," and in the next paragraph refers to "the three fused ocelli of Limulus" (p. 344). 



