ROSÉN, STUDIES ON THE PLECTOGNATHS. 13 



and the Tetrodontids live among coral reefs and in suchplaces 

 as bights etc. where they may easily be cut off from a larger 

 amount of water or perhaps be left qiiite dry when the tide 

 falls. From this point of view it is not surprising, that there 

 is not an air-sac developed for the purpose of indirect respi- 

 ration in the Molids and the Ostraciontids. The former are 

 pelagic in habit and the latter live on the bottom of the sea, 

 though in rather shallow water. It is well known that some 

 fishes take air in the alimentary canal for a direct respiratory 

 purpose. The question may therefore be proposed whether 

 in the Plectognaths direct respiration takes place in the air- 

 sac or not. A glance at fig. 1 on Pl. I, which illustates a 

 section through the air-sac of Spheroides would perhaps at 

 first seem to support such an opinion. The inner surface 

 of the air-sac is covered with a great number of large folds. 

 But as I have mentioned above, these folds are perfectly 

 smoothed out when the sac is inflated. The sections that I 

 have made through such specimens have shown this clearly. 

 Also the arrangement of the vessels does not speak for a 

 direct respiration. The air-sac is supplied with blood by 

 the gastric artery. It receives arterialised blood. This 

 latter circumstance is, of course, no absolute proof against 

 the view of a direct respiration, in that case the respiration 

 by means of the air-sac would of course only be of a temporary 

 importance when the blood does not get arterialised enough 

 in the gills. But the absence of an enlarged surface and the 

 fact that the walls of the air-sac are not especially rich on 

 blood-wessels are sufficient proofs. 



Some remarks may be given on the mode in which the 

 air-sac is filled and emptied. The air is of course taken in 

 by means of swaDowing movements. In the forms in which 

 the air-sac is represented only by a dilated stomach the air 

 is retained here, as Thilo has found, by a pyloric valve and 

 by means of circular muscles in the oesophageal wall. The 

 air is, of course, repulsed by a relaxing of the last named 

 muscles and perhaps to some extent also by action of the 

 body muscles. In the Diodontids and Tetrodontids, I have 

 found an arrangement of the ventral body muscles, that is 

 in all probability directly adapted for the purpose of expelling 

 the air. As mentioned above there is in these forms an in- 

 teriör layer of longitudinal and a thin exteriör layer of träns- 



