10 



ARKIV FÖR ZOOLOGI. BAND 7. NIO 30. 



with the body wall biit a closer examination shows clearly 

 that we have the same condition as in D. hystrix; that the 

 air-sac is very much enlarged and the chords very short so 

 that the air-sac ahnost touches the ventral body wall. In 

 an adiilt specimen of Chilomycterus geometriens, a genus allied 

 to Diodoii, I found the condition to be the same as in D. 

 holacanthus . 



cU 



f>e\ 



IT U I Ii 1 I I . 11 1 11] I I llVi r j 



Fig. 6. Diodon hystrix L. et. connective tissue of the chords uniting the 



air-sac with the ventral body wall; du duodenum; Zmlongitudinal muscular 



layer; oes oesophagas; per peritoneum; sk skin; st stomach; trm träns - 



verse muscular layer. 



That the air-sac in the Plectognaths forms a part of the 

 alimentary canal, is clearly shown above. All previous authors 

 seem also to have had the same opinion, except Sörensen, 

 who has given a quite unintelligible explanation, considering 

 it to be a sac-shaped duplicature of a subcutaneous layer of 

 connective tissue, which is not present in any other fishes. 

 I do not think that a criticism of this view^ is necessary after 

 the inquiry on the air-sac given above. And the same is 

 the case with the discussion about the homology of the air- 

 sac with the lungs, into which several earlier authors have 

 cntered, and on which v. Baer has given a summary (p. 47). 

 The homology with the lungs w^as låter on again maintained 

 by Albrecht. w^hose paper I have not seen. What part of 

 the alimentary canal does the air-sac represent? I have al- 



