432 RECTAL GILLS IN LARViE OF ANISOPTERID DRAGONPLIES, 



If, on the other hand, (as seems more probable) the proportion 

 be less, oxygen will pass in to a like extent. Or if, as may well 

 be the case, the proportion is the same from the start (seeing 

 that both egg and larva are, throughout the operation, immersed 

 in water), equilibrium will be established from the start. 



71ie Process of Diffusion during larval life. — Whatever be the 

 composition of the original gas in the tracheae, it is the oxygen 

 in it which is required for the metabolism of the growing larva. 

 There is, therefore, from the time that equilibrium is established, 

 a continuous drain on the oxygen in the tubes. This gas is 

 taken up at all points of the body by the cells of the developing 

 organs, hy direct tracheal supfly. Thus there is a steady diminu- 

 tion of the partial pressure of oxygen going on on one side of the 

 diffusion-membrane, while the partial pressure of the same gas 

 on the other side of the membrane (i.e., in the water) remains 

 constant. Hence there must be a tendency for oxygen to diffuse 

 from the water into the tracheal system at all jioints where the. 

 two are brought into contact. The problem is the same here for 

 the integument as for gills placed in any position on the body. 

 What we have now to see clearly is, that the suitable conditions 

 are only set up, in our larvae, to any appreciable extent, in the 

 case of the rectal gills. In Fig. 6, I offer a comparison between 

 {a) the ordinary body-wall, (h) the undifferentiated rectal wall, 

 and (c) the gill-wall of our larva. In (a), the cuticle is thick and 

 many-layered, the hypodermis is also fairly thick, and composed of 

 separate cells, and there is no special arrangement of tracheae for 

 the reception of the diffused gas (supposing that any were to 

 pass through). In (6), the conditions are more favourable, in so 

 far that the cuticle (intima) is thinner and single-layered, the 

 epithelial cells flatter (though still separate), and the tracheal 

 supply, on the whole, richer and finer than in the case of (i). In 

 (c), finally, not only are the cuticle and epithelium reduced to 

 the finest possible dimensions, and the cell-boundaries of the 

 latter obliterated, but thousands of minute capillaries are actually 

 sunk into the epithelial syncytium, so that they can receive 

 directly the supply of diffusing gas. Could any arrangement 

 possibly be simpler, or more perfect for the action of diffusion ? 



