neal: nervous system in squalus acanthias. 267 



I regard the mouth of Amphioxus as homologous with the left half of 



the mouth of Crauiota aud the club-shaped glaud as its autimere. That 

 the mouth of Amphioxus as au organ of the left side is exactly homolo- 

 gous with the left half of the mouth of Squalus appears to me probable 

 on the a priori ground that it is improbable that an organ of the same 

 function should be twice acquired in the Vertebrate series ; and also be- 

 cause the region of fusion of endoderm and ectoderm to form the mouth 

 cleft is in both these forms ventral to the constrictions which separate 

 the second ^ud third mesodermic segments (1st and 2d myotomes). 

 The club-shaped gland also appears as an entodermic diverticulum below 

 the constriction between the second and third mesodermic segments 

 of the right side, that is, opposite the mouth diverticulum, and I there- 

 fore, in agreement with van Wijhe ('93), regard it as the autimeric 

 gill cleft.^ 



In the fourth segmeiit the following points of resemblance are to be 

 noted. Somatic musculature and a somatic ventral nerve are present. 

 While in Squalus the pair of visceral clefts which bounded anteriorly 

 the splanchnic portion of this segment have disappeared, leaving no trace 

 beliind except in the neuromere with which they were connected, in 

 Amphioxus only the right visceral cleft has been thus lost. The left 

 visceral cleft, however, disappears ontogenetically without leaving a trace 

 behind it. A further diiTerence in the two forms appears in the fact that, 

 whereas in Squalus the dorsal nerve has disappeared (or fused with the 

 trigeminus), the dorsal nerve of the left side in Amphioxus is the first 

 of the nerves which innervate the musculature of the velum (van 

 Wijhe). 



With the fifth segment in both forms begin the permanent visceral 

 clefts. In agreement with Willey ('94), I regard the first secondary cleft 

 as autimeric to the second primary cleft. Their fusion with the ecto- 

 derm below the mesodermic constriction between mesoderm segments 

 4 and 5 (myotomes 3 and 4) is the evidence for their relation to 

 this, the fifth segment. I therefore consider the first pair of permanent 

 visceral clefts in Amphioxus as the exact homologues of the hyomandibu- 

 lur clefts of higher Vertebrates. As has already been stated by Willey 

 ('94), all except eight of the primary clefts (starred in the table), which 

 become paired with eight autimeric clefts, undergo atrophy. In conse- 



1 Willey ('94) gives reasons for regarding the club-sliaped gland as the antimere 

 of the first primary visceral cleft. His reasons are based on topographic relations 

 in stages when the primitive topographic relations are considerably changed, and 

 they seem to me less strong than the reasons stated by van Wijhe and myself. 



