ROSÉN, STUDIES ON THE PLECTOGNATHS. 15 



not yet been developed, thus quite independent of the infla- 

 ting-power, but these modifications have, on the other hand, 

 been of importance for this organ rendering its development 

 possible. This secondary relation between the air-sac and a 

 certain anatomical structure thus attained, the f unction of the 

 first mentioned organ perpetually increasing in importance, 

 has at last also directly influenced the organ in question. 

 Though the detailed account of the structure of the organs 

 which has become modified to any extent in connection with 

 the development of an air-sac, will be given in the respective 

 sections of this series of studies, I shall give here already a 

 summary of these modifications and discuss whether they are 

 directly caused by the peculiar power of inflating the belly, 

 or only secondary to it. 



The anatomical system that ought first to be taken into 

 consideration from this point of view is the alimentary canal. 

 From this the air-sac has taken its origin, as shown above. 

 It would, perhaps, seem probable that an organ, the function 

 of which is of a digestive nature, when taken into a respira- 

 tory service, though only a secondary one, should be somewhat 

 altered. But this is nolr the case, except, of course, the very 

 part of the alimentary canal, that is developed into an air- 

 sac. The absence of glands in the stomach may not be re- 

 garded as caused by the dilatation into an air-sac. Even in 

 Balistes, where the stomach forms no such sac, the stomach 

 glands are wantig (Edinger), and the same condition is found 

 in several other Teleosts which possess no air-sac. 



The most important and interesting alterations we find 

 in the sheleton. Referring to the description of this organ, 

 which I shall give in a special section, I shall here confine 

 myself to a statement of the most important modifications. 

 It is evident that an inflation of the belly like that we have 

 seen to be the condition in the Diodontids and Tetrodontids 

 would not be possible in forms which possess well developed 

 pelvic bones. Thilo has made interesting researches on the 

 different forms that the pelvic bones show, to which paper 

 I refer, adding some remarks. The pelvic bones ofeachside 

 have coalesced forming a large piece, the end of which is 

 provided with a large process (Thilo's s>Bauchstacheb) and 

 with a ventral spine. In Balistes these bones form, as is 

 shown in Fig 3, the ventral wall of the body cavity. They 



