290 



THE AMERICAN NATURALIST [Vol. XLI 



by watching the movement of sohd particles suspended in the 

 water in the vicinity of the mouth. Much the greater volume of 

 water, however, appears to enter through the spiracle.^ During 

 an expiration the spiracle is shut, while the mouth tends to close 

 but does not close tightly. The mouth action was always a little 

 sluggish as compared with the action of the spiracle, especially 

 in opening. As nearly as I could determine, mouth and spiracle 

 closed together, but the opening of the mouth was slightly later 

 than that of the spiracle. As spiracle and mouth close, the bran- 

 chial region is contracted and the water contained in the gill 

 chambers is forced out through the gill clefts. 



That water does not flow out through the mouth as well as 

 through the gill clefts during an expiration is probably due to the 

 action of a well developed respiratory valve similar to those de- 

 scribed for teleosts by Dahlgren ('99). The dorsal flap of the 

 valve (Figures 1 and 3, vlv. d.) is a conspicuous bilobed fold of the 

 oral membrane, while the ventral or mandibular flap — a less 

 extensive fold — is broadest in the median region of the lower jaw 

 and becomes much narrower towards the sides of the mouth. 

 Judging from the relative widths and the positions of the two parts 

 of the valve, it appears that the prevention of outflow through the 

 mouth must depend mainly upon the action of the dorsal flap. 

 Garman ('74) mentions only the dorsal one of these two folds. 



The elevation of the forward end of the fish above the surface 

 on which it rests would seem to facilitate the respiratory process. 



flat-bottomed glass vessel. The vessel was placed upon a higlf table so^that 

 one end projected some distance beyond the edge of the table. The fish was 

 induced to lie with its head in the overhanging part of the vessel. I found that 

 an object held just underneath the mouth could be seen directly through the 

 head of the fish by looking, at the proper angle, into the spiracle as it opened, 

 and, similarly, an object held just above the spiracle could be seen by looking 

 upward into the mouth as it opened. This was sufficient proof that mouth 

 and spiracle were open at the same time. In order to see the dorsal and ven- 



total reflection from the siirfinc of the wntcr H;i\in>: tlir \\;itci- :ii ;i ccrtMiu 

 depth and looking upward Irom uinlcrn(';ii h the ii\ crluiiiiriiii: xc-^rl at ]\\<\ 



sufficient clearness, while at tli<" same time 1 had a direct view ot the \ciitral 

 surface of the head. 



