110 EDWARD PHELPS ALLTS jr., 



of its lips from before backward; and then the open chariüel that 

 forms the posterior part of the opening in larvae either becomes con- 

 verted into a closed canal by a continuation of the same process, or 

 it is ])ushed backward and upward onto the anterior surface of the 

 first branchial cleft. Diifereut stages in this process are readily seen 

 in larvae of different ages. In other words, the pseudobranchial 

 chamber, which, in 12 mm larvae, opens directly inta the mouth 

 cavity by a sht-like opening which extends the full length of the 

 pseudobranch, and, posteriorly, is continuous with an open groove 

 which opens partly into the mouth cavity and partly into the first 

 branchial cleft, opens, in the adult, into the anterior end of the first 

 branchial cleft alone, by a rounded or oval opening. Whether the 

 open groove in 12 mm specimens actually represents a part of the 

 spiracular cleft, or not, I have not attempted to investigate. I treat 

 it as such, as others have done before me. 



The pseudobranchial chamber, in 12 mm larvae, is a long and 

 tall, but narrow space which extends upward and laterally along the 

 lateral surface of the skull, at an angle of about 45'^ to the general 

 horizontal level of the dorsal surface of the mouth cavity in its vicinity. 

 In it the limits of the two parts described by Wright as the pseudo- 

 branchial chamber and the diverticulum of that chamber, or spiracular 

 cleft properly so-called, can not be easily distinguished. What is 

 probably the chamber, as defined by him, has its widest portion im- 

 mediately dorsal to the pseudobranch, and is wider posteriorly than 

 anteriorly. Its anterior end extends forward shghtly beyond the an- 

 terior end of the pseudobranch, and also slightly beyond the anterior 

 end of the slit-like aperture that leads from the chamber into the 

 mouth cavity. Posteriorly the chamber slopes gradually downward 

 and backward to the anterior edge of the adductor hyomandibularis 

 muscle, which marks its posterior limit. The dorsal diverticulum is 

 but slightly developed, at this age, as a perceptible cleft, but a cord 

 of sensory and other tissue, in which the diverticular cleft will later 

 develop, projects upward and laterally from the pseudobranchial 

 chamber somewhat anterior to its middle point. In later stages, where 

 the diverticulum, as a cleft, is more developed, the sensory part of 

 this cord of tissue, which forms the spiracular sense organ of Wright's 

 descriptions, is seen to lie along the anterior edge of the diverticulum, 

 and along its postero-mesial wall posterior to its anterior edge. In 

 12 mm larvae this cord or line of tissue lies anterior to the dorsal 

 end of the hyomandibular, and in the specimen cut in transverse 



