198 EDWARD PHELPS ALLIS, JUN. 



tliat arises from the ventral division of the nerve of the arch, 

 wholly separate and apart from the ramus dorsalis of the 

 segment. The innervation of the organ in Mustelus by a 

 branch of the ramus oticus, usually considered as a ramus 

 dorsalis, and its position in what Hoffmann considers as a 

 remnant of the mandibular cleft, are accordingly both de- 

 cidedly opposed to the assumption that the organ finds its 

 homologue in the hypoblasiic organs of Ammocoetes. Hoff- 

 mann avoids this difficulty, in so far as it relates to the 

 innervation, by the very legitimate conclusion that both the 

 ramus oticus and the ramus buccalis, the nerve from which 

 the oticus arises, are ventral and not dorsal nerves. 



In my work on Amia (2) I was led to suggest that the 

 spiracular organ of that fish was simply an infraorbital lateral 

 canal organ, of ectodermal origin, that had wandered into the 

 spiracular cleft as the external opening of that cleft was 

 closed. If the same assumption be made regarding Mustelus, 

 it is evident that, if Hoffmann's conclusions are correct, the 

 cleft into which the organ wanders must be the mandibular 

 and not the spiracular cleft. It is also probable, under 

 the same assumption, that another infraorbital organ similarly 

 wanders into the next anterior, or labial cleft. That these 

 assumptions are not improbable is evident from Hoffmann's 

 account (36, p. 339) of the similar wandering of what he 

 calls rudimentary " Hautsinnesorgane '' from the outer sur- 

 face into the branchial clefts related to the glossopharyn- 

 geus and vagus nerves. These rudimentary cutaneous sensory 

 organs are said to be formed in relation to the ganglion of the 

 ramus ventralis of the nerve concerned, and they are said to 

 abort completely soon after entering the related cleft. From 

 Hoffmann's account of them they seem not to be considered 

 by him as organs homologous with the organs of the 

 lateral canals, but it seems probable that they find their exact 

 serial homologues in the so-called spiracular organs of 

 Mustelus, excepting only in their being innervated by a 

 branch of a ramus ventralis instead of by a branch of a so- 

 called ramus dorsalis. This difficulty is, however, wholly 



