ROSÉN, STUDIES ON THE PLECTOGNATHS. 9 



Diodoiitidae. A very interesting form of the air-sac is 

 present in the genus Diodon. In an adult specimen of D. 

 hystrix L. I found a liighly enlarged stomach from the ante- 

 rior and ventral wall of which narrow chords arise uniting 

 it with the abdominal wall (Fig.^ 6; PL II, fig. 2). These 

 chords consist of connective tissue, without any muscular or 

 elastic fibres. Their origin from the stomach is generally 

 very broad, and the base of several chords are often united 

 forming large flaps. Their attachment to the ventral body 

 wall is slightly enlarged (PL II, fig. 2). The peritoneum 

 must be very peculiarly developed in this species. The pre- 

 servation of the specimens has not allowed me to find the 

 epithelial layer of the chords, but as the peritoneal layer con- 

 sists of flattened cells there is no wonder that it is not pre- 

 served. We have every reason to suppose that the peritoneal 

 layer lines the surface of all these chords as shown in Fig. 

 6. Underneath the peritoneal layer of the ventral body wall 

 there is a thin stratum of connective tissue, exteriorly to which 

 runs a layer of longitudinal and transverse muscles as in 

 Spheroides. There is no fold which forms a boundary between 

 the stomach (s. str.) and the air-sac as in Spheroides. Some 

 of the earlier authors seem sometimes to have mistaken the 

 air-bladder for the air-sac and vice versa. Some of them have 

 described the air-sac in Diodon as »tres celluleuse». But all 

 these old descriptions contain only a few words and are never 

 accompanied by any figures, so it is quite impossible to get 

 a clear idea about the form of the air-sac in the respective 

 species. Perhaps the chamber between the chords, which is 

 nothing but the common body cavity, is sometimes mistaken 

 for an air-sac »tres celluleuse». But to try to find out the 

 meanings of these earlier descriptions would be of no use. 

 From a genetical point of view the condition I have described 

 above for Diodon hystrix is of a great importance. It forms 

 an interesting stage between all the species in which the air- 

 sac represents only a dilated stomach and Spheroides and 

 Tetrodon (except T. rubripes and others?), in which the ventral 

 wall of the air-sac has perfectly coalesced with the ventral 

 body wall. A stage a little further advanced in the direction 

 of Spheroides and Tetrodon, seems to be present in D. hola- 

 canthus. I have studied only small specimens of this species. 

 At a hasty glance it would seem as if the air-sac were coalesced 



