2 ARKIV FÖR ZOOLOGI. BAND 7. NiO 30. 



chief part of the specimens on which these studies are based 

 was collected by myself during a visit to the Bahamas. 

 Unfortunately, I had no opportunity of making any experi- 

 ments on the fiinction af the air-sac. The anatomy has been 

 studied not only by dissecting alcohol or formol specimens, 

 but as regards Spheroides testudineus L. also in tvo series of 

 microscopical slides of young specimens, stained with v. Gie- 

 son's mixture or with boraxcarmin, Bismarckbrown and in- 

 digcarmin and in sections of different parts of the alimentary 

 canal, stained with haematoxylin, v. Gieson's mixture, kre- 

 sofuchsin and other staining fluids. 



I. The morpholog-y of the Air-sac. 



An air-sac is not developed in all Plectognaths. In the 

 species which are provided with this organ, it always forms 

 a dilatation of the anterior part of the alimentary canal. A 

 study of its shape will show a continuous series of different 

 stages of specialization from a hardly traceable dilatation to 

 a highly developed air-sac, which has become united with the 

 ventral body wall. I shall describe, below, the air-sac in the 

 species which have been studied by myself or by other authors 

 including also an account of the corresponding part of the 

 alimentary canal in the species without an air-sac. 



Molidae. I have not examined any species belonging 

 to this family, but all the authors that have written about 

 them agree in stating that the air-sac is absent. According 

 to Jacob, Lutken and Steenstrup the stomach is very small, 

 hardly distinct from the other part of the intestine. No dila- 

 tation at all is to be found. 



Ostraciontidae. I have studied of this family Lactophrys 

 trigonus L. (Fig. 1). The stomach is distinguished from the 

 duodenum' by a feeble fold but there is no well marked boun- 

 dary between the stomach and the oesophagus. It is quite 

 evident that forms like the Ostraciontidae (»Trunk-fishes»), 

 which all have the body encased in a carapace formed by 

 bony plates, are not able to inflate the belly; thus we cannot 

 expect to find an air-sac for this purpose. But, as will be 

 shown låter on, the function of the air-sac in its most pri- 

 mitive stages seems to be of a respiratory nature in that it 



