264 bulletin: museum of comparative zoology. 



bands force their way. In this way a special anterior portion of the 

 body cavity, prseoral and praeenteric, is produced, and is at first com- 

 pletely filled by a compact mass of rounded cells derived from the meso- 

 dermic bands. . . . The anterior, or prseoral portion of the body cavity, 

 of which we have just traced the origin, is, and subsequently becomes in 

 a still more pronounced way, the cavity of the snout, or prceoral lobe." 



On the other hand, the preoral "head cavities" of Amphioxus, whicli 

 Willey homologizes with the "preoral lobe" of Ascidian larvae, are formed, 

 as stated by Hatschek ('81), from an abstricted portion of the preoral 

 archenteron. The differences in the formation of these two structures, 

 therefore, seem too striking to permit their being considered completely 

 homologous with each other. 



Evidence has already been given in this paper which, in my opinion, 

 makes it impossible to homologize the preoral "head cavities" (anterior 

 entodermic diverticula) of Amphioxus with the " premandibular head 

 cavities " of Craniota. The morphologically anterior portion of the arch- 

 enteron, the " anterior head cavities " (Piatt), are the only structures 

 in higher Vertebrates which, in my opinion, can be homologized with 

 the "head cavities" of Amphioxus. Homologizing, therefore, the "an- 

 terior head cavities " in these two forms, I subuiit on the opposite page 

 for comparison with Squalus the table of the anterior eleven segments 

 in Amphioxus as I interpret them. 



If we compare Tables III. (p. 253) and V., we find the following 

 fundamental resemblances in the segments of Squalus and Amphioxus. 



Of the component elements- of the first segment, that which I have 

 regarded as the somatic element, consists of paired cavities cut off from 

 the anterior portion of the archenteron. Since in both cases these 

 cavities represent ventral as well as dorsal and lateral portions of the 

 archenteron, it is impossible to contend that they contain only the 

 mesodermic element of the segment. It seems not improbable that 

 potentially they represent also the visceral-pouch element between this 

 and the following segment. The opening of the left of these in Amphi- 

 oxus to the exterior as the preoral pit may be regarded as evidence 

 favoring this view. Moreover, M. Fiirbriuger ('97, p. 633) finds a late 

 diff"erentiated and rudimentary myotome, which lies anterior to the 

 dorsal paired nerves II, which would, if present as stated by him, repre- 

 sent the mesodermic element of this segment, and the " anterior head 

 cavities " would in consequence necessarily be regarded as modified or 

 abortive visceral pouches, as held by Kupffer. Since my sections of 

 Amphioxus give me no evidence of this rudimentary myotome, I hold 



