SECEETORY SACS IN TELEOSTEI. 551 



physiological significance of the posterior nostrils similar to that 

 of the internal nares are absent. 



The walls of the posterior nostrils, again, act as valves whose 

 function is to let water pass from within outwards. In the free- 

 swimming forms, therefore, when both nostrils are present, it is 

 probable that the movements of the fish through the water 

 suffice to induce the passage of water through the nasal cavity ; 

 and this flow is controlled by the anterior nostril, whose 

 walls may be prolonged into a contractile tube or flap-like 

 covering. Where only one external nostril is present, it is the 

 anterior which must carry on the functions of the two of the 

 other forms. And hence, if analogies are to be drawn, the 

 anterior nostrils might be likened physiologically to the internal 

 nares, and the posterior nostrils to the gill-openings. Similarly, 

 when nasal sacs are present the displacement of the sacs ensures 

 the circulation of water through the nasal cavity, just as the 

 movements of the mouth, and gill-covers cause the passage of 

 water through the mouth, and the posterior nostrils are again 

 only comparable to the gill-openings. 



We may proceed now to a consideration of the other water- 

 breathing Vertebrates in which the function of respiration is in 

 part carried on by a naso-pharyngeal communication. In the 

 Cyclostomes we find an analogy to what is seen in Gynoglossus. 

 In the Petromyzontidae there is a prolongation from the nasal 

 cavity backward to the roof of the mouth ; in the Mysinidae the 

 latter is pierced, and the communication thus opened is supposed 

 to fulfil a respiratory function. The origin and development of 

 this so-called prolongation of the nasal organ bave been described 

 by Dohrn * for Petromyzon, and recently somewhat briefly by 

 Dean f for Bdellostoma. 



Leaving aside the apparent remarkable differences alleged by 

 Dean in the development of these two forms, it is clear that the 

 internal nares of the Myxinidse are formed by the communication 

 of the hj-pophysis with the gut. According to Dean the hypo- 

 physis arises in Bdellostoma before the external openings of the 



* Dohrn, A. : " Studieu zur Urgeschichte des Wirbeltliierkorpers. III. Die 

 Enstehung u. Bedeutung der Hypophysis bei Petromyzon Planeri" Mitt. Zool. 

 Stat, zu Neapel, 1883. Vide Howes, Gr. B. : " On the Affinities, Inter-relation- 

 ships, and Systematic Position of the Marsipobranchii," Tr. Biol. Soc. 

 Liverpool, vol. vi., 1891, p. 122. 



t Dean (Bash ford) : ' On the Embryology of 'Petromyzon StoutV Kupffer's 

 Festschrift, Jena, 1899, p. 269. 



