GAR-PIKES, OLD AXD YOUNG. 



11 



of the ventral surface. With two African Ganoids, Polypterits and 

 Galamoichthys (as also stated by Miiller, and verified by the writer as 

 to the latter genus) the sac is double, and communicates with the ven- 

 tral side in the median line ; but it is slightly cellular, as in Meno- 

 branclius. 



A. sturgeons and many Teleosts. 



B. Amia and Lepidosteus. 



C. ErythrinuK. 



G. Ceratodus. 



D. Polypterus, Calamokhihys. 



E. Lepidosiren, Protopterus. 



F. Eeptiles, birds, mammals. 



Fig. 7. Diagbams bepresenting the Con~nection' betwe'^n the Air-bi^adber or Lxing and 

 THE Alimentary Canal in Certain Vertebrates. 



^/, the alimentary canal, vl, the air-bladder. J.Z>. the air-diict. 



The flsjures at the right show the alimentary canal and air-bladder from the left side ; those at the 

 left represent cross-sectioiis more or less foreshortened in some cases. 



A A' represent the simple condition-connections of the air-bladder in the stiirgeons {Acipen- 

 ser) and in most Teleosts where the air duct remains open. B B' represent the arrange- 

 ment in Amia and Lepidosteuf!. where the duct opens upon the dorsal side of the throat, but 

 the bladder is more or less cellular. The hinder end of the bladder is left open to indicate its 

 great length in Lettidoxteus. In C C is shown the arrangement in Ei-ythrinus. The bladder 

 is still upon the dorsal side, but the front part is separated from the hinder two-thirds by a 

 constriction, and the long duct passes f irward from just behind the constriction to enter the 

 left side of the throat. There are fibrous partitions in part of the bladder, but I do not know 

 that they are vascular. The condition in Ceratodus is shown at G ; the bladder is single but 

 vascular, and the duct opens on the ventral side, but not in the middle line. 



In the remaining figures the air-duct opens on the lower or ventral surtoce of the throat, and 

 the air-bladder is in two parts, which unite at the duct, but separate backward and lie 

 upon the sides of the stomach, or even to gome extent upon its dorsal surface nest the back- 

 bone. In the side-views only the left sac is seen ; in the cross-sections the whole is fore- 

 shortened so as to bring it into one plane. In Polypterus and Calamoichthys the inner surface 

 of the sacs is nearly smooth, but in Lepidosiren, as in the salamanders, it is more or less 

 folded and vascular, and is also connected with the heart bv special vessels. In the reptiles, 

 birds, and mammals, the duct or trachea soon divides into the two bronchial tubes. 



Finally, in the " mud-fishes " of Africa and South America (Pro- 

 topterus and Lepidosireii) the duct is ventral, and the air-bladder is a 

 double and lung-like sac with stiff walls. 



This series seems to connect the air-bladder of the fishes with the 

 lungs of the true aerial vertebrates, and to remove the objection 



