430 



RESPIRATION OF MESCX.OM rjl I \s. 



that this is true for all species. Figure 6 sliovvs the ifolds and 

 their tracheation, and Figure 7 a l)it of the margin of a fold, 

 with its tracheal loops. In Figure 5 the buttress folds appeal-. 



In the base of each buttress fold, where it joins the longi- 

 tudinal fold, is a bit of fatty tissue, F. This feature 1 liave 

 found in all genera studied by me, save one. Likewise, there is 

 on the cei)halic side of the gill a tough "cushion," (". of thick 

 epithelium. This also is a constant feature. I am not aware 

 that the function of these parts is known. 



The buttress folds are not set at right angles to the longi- 

 tudinal folds, but are skewed cephalad. They hang in the 

 rectum with their points directed cephalad. 



i'li. 



The respiratory rectum, with its delicate i>arts. is not freeh' 

 open to the water at all times. Hidden under the five terminal 

 appendages of the abdomen is a valve oif three membranous 

 folds, which may be closed at will by the animal. I have seen 

 the nymi)h of LiheUula open and close this valve coiuinucnisly 

 and rapidly when in water with particles of dirt in it. 



The respiratory rectum is connected in Mesogoiuphns as in 

 all Aeshnidre, with the anus by a canal. AC. occupyiui;' the tenth 

 abdominal somite. In the sixth somite a broad band of muscle 

 passes over the intestine, and appears to act as a valve to keep 

 the water oiu of the digestive part of the alimentar\- canal. In 

 specimens of various genera I have found thi^ band c]ami)e(l 

 down hard on to the intestine. 



