IX] THE RESPIRATORY SYSTEM 189 



(5) The hypobranchial tissue (lib}. The space at the base of 

 the gill, between the two walls, is more or less completely filled 

 with a peculiar tissue surrounding the larger tracheae. This tissue 

 consists of large rounded lobules, each with one or more rounded 

 nuclei. The cell-boundaries are not very distinct. It has been 

 called "connective tissue," and also "adipose tissue," but most 

 resembles the latter. It appears to be a specialized development 

 from the fat-body [145, 1793. 



5. Physiology of Respiration. 



We cannot here enter into the details of the extremely prob- 

 lematical and difficult subject of insect respiration in general, 

 and aquatic larvae in particular. No really acceptable explanation 

 of the action of the rectal gills has yet been put forward. The 

 solution, however, appears to he along the simple line of ex- 

 planation indicated by Bis [138], following Lowne. 



It seems clear that the primary factor in rectal respiration 

 is the epithelial syncytium, and that its action is exceedingly simple. 

 The presence of a layer of separate cells, in the unspecialized 

 rectum, formed, I believe, a definite barrier to the rapid absorption 

 of oxygen, when the larvae first took to the water. By fusion of 

 the cells into a single protoplasmic syncytium, and by extensive 

 stretching of the same to the necessary tenuity, there is developed 

 between the circulating water and the underlying capillaries a 

 structure just strong enough to prevent injury to the latter, but 

 sufficiently permeable to act as a diffusion membrane. The action 

 of such a membrane would be as follows : 



When the larva hatches, the tracheal system is quickly filled 

 with carbonic acid gas, derived from the region of the mid-gut, 

 as already described on p. 70. By this means, all the capillary 

 loops become filled with gas. As soon as this is accomplished, 

 water is drawn into the rectum, and regular respiratory move- 

 ments begin. At the start, the pressure of C0 2 in the capillaries 

 is about 760 mm. In the water drawn into the rectum, the 

 partial pressure of this gas is less than 1 mm., while that of 

 is about 160 mm., and that of N about 600mm. It follows, 

 therefore, that C0 2 must diffuse rapidly outwards from the 

 capillaries into the water, and that it will be replaced, in the 



