STSTEM OF SACCOBRANCnUS POSSILIS. 51 



fourth on both sides is considerably larger than the others, and, 

 after coursing along the fourth gill-arcb, is continued upon the 

 ventral wall of its respective air-sac. The first, second, and third 

 go to their several gills in the ordinary way. The first on the 

 right does not differ in size from its companion on the left, and 

 rapidly diminishes in calibre in its course along the gill, so that 

 I was unable to trace it more than half an inch or so. 



Unfortunately it was not possible to inject this fish ; but the 

 vessels were sufficiently conspicuous to leave no doubt in my 

 mind as to the accuracy of this observation. 



It is to be observed that Hyrtl made his observations upon 

 Saccohranclms singio, so that it is possible that this distribution 

 of the branchial vessels may have a specific significance, and not 

 be merely a case of individual variation. 



"With reference to this arrangement of the aortic arches in 

 SaccoirancJius, it is interesting to briefly review the work that 

 has already been done in connection with the blood-supply to 

 organs of aerial respiration in fishes and the higher Vertebrata. 



Beginning at the top of the scale and working downwards, we 

 find that a general law has been laid down by Boas * to the eff'ect 

 that in the Amphibia and all higher Yertebrata the pulmonary 

 artery is always derived from the fourth branchial aortic arch f. 



This generalization is considerably strengthened by van Bem- 

 meln's % discovery in embryonic E-eptilia and Aves of two gill- 

 clefts and an aortic arch lying between the systemic and pulmonary 

 arteries ; and still further by Zimmermann's § demonstration of 

 an aortic arch in the same position in embryos of the rabbit 

 and Man. 



Coming now to the Dipnoi, amphibious fishes whose swim- 

 bladder has been modified for purposes of aerial respiration, 

 matters become complicated by the reduction and compression of 

 the branchial apparatus. It is possible, however, in Ceratodus, 



* Boas, Morph. Jahrb. Ed. vii. 1882, p. 488, " TJeber den Conus Arteriosiia 

 und die Arterienbogen der Amphibien ; " and Morpb. Jabrb. Bd. xiii. 1887-88, 

 p. 115, " Ueber die Arterienbogen der Wirbelthiere." 



t That is, the 6tb -visceral aortic arch. For simplicity's sake I count from 

 tbe 1st branchial aortic arch. 



:j: Van Bemmeln, "Die Visceraltascben und Aortenbogen bei Eeptilien und 

 Vogeln," Zool. Anzeig. 1886, pp. 528 & 543. 



§ Zimmermann, " Ueber einen zwiscben Aorten- und Pulmonatbogen gele- 

 genen Kiemenarterienbogen beim Kaninchen," Anatomiseh. Anzeig. 1889, 

 p. 720. 



