ON THE DEVELOPMENT OF LIMULUS LOXGISPIXA. 9-^ 



But none of these points so happen in the development of the pulmonary 

 sac of the Arachnida. Hence I think the pidmonary e-ic of the 

 Arachnida has been produced by the subsidence of the lamelligerous 

 appendage of a Limulus-Wke ancestor. 



Tlie number nnd position of the eyes are exactly the same in 

 Liiimhis and in tlie Choidata. The median eyes of Lhiiiihis are repre- 

 sented by tlie pineal eve of the C'hordata, which is pioi):!l)Jy frrmed by 

 the union of a pair of median eyes such as tliose in Tj'mnliis, The 

 lateral eyes of Liniiilns isre repre>erjted by the paired e}es in the 

 Chordata. The paired eyes of the Chordjit;i differ from the lateral eyes 

 of X(?'»«t/z(s by having- their retinal portions inverted; but this is the 

 effect of the closure of the medullary plate with the optic area into a 

 tube. The retina of tlie [)aired eyes of the C'hordata is also composed 

 of ommatidia — a cone surrounded by many rods. 



The lateral eyes of Liimdus receive their nerves from the pre-oral 

 cephalic lobe. Then the ancestral mouth of the Chordata must be 

 searched for^behind the optic chiasma, and witliin tlie ring of the ner 

 vous system. We have an invagination in the nervous system, which 

 cannot be explained otherwise than by considering it as the degenerated 

 stomodai'um of the ancestor. It i-^ the infundibulum. The view that 

 the infundil)ulum is the remnant of the ancestral mouth is maintained 

 by Cunningham, l^itten, Gaskell, &c. The facts tliat the infundibulum 

 is closely in contact with the anterior end of the mesenteron in the 

 embryonic stage, and that it is behind tlie optic chiasma, fixvours 

 this conclusion. 



The origin of the mesolilast, separation of the ner\'e-cord from the 

 epiblast, formation of the lieart, the number and position of the eyes, 

 the grooves in the Ijrain, &c., make Limuliis approach in(_)re cl(3sely to 

 the Chordata than it was thought to do formerly, and accordingly make 

 the Arthropoda more nearly related to the C'hordata than to tlie Vermes. 

 Tokyo, June, 19, 1891. 



