THE LAHVA OF TIPULA MAXIMA. 129 



through these gills, flowing- outwards along one channel and returning along the other, 

 whence it passes dorsalwards to enter the dorsal vessel through its posterior opeuing. 

 The now through these gills can be distinctly observed in the young larva, when it 

 is seen to be intermittent, following the pulsation of the dorsal vessel. 



Blood-gills are also described by Fr. Miiller as occurring in larvae of Trichoptera as 

 eversible finger-like tubules, into which the blood flows. They do not generally possess 

 a tracheal network, but occasionally very tine tracheal branches enter. Pictet describes 

 them in pupa? of Caddis-flies, and Schiodte in the larva of Pelobius. They are 

 present in Chironomus and some other larva?. 



Blood-gills differ from tracheal gills found in many aquatic insect-larva? in the 

 character of the tracheal supply and in the presence of large blood-sinuses. In the case 

 of tracheal gills probably the exchange occurs directly by diffusion from the water 

 to the trachea?, while in the case of blood-gills the blood itself plays an important part. 

 G. Gilson, in describing the rectal blood-gills of Odonata (Journ. Linn. Soc, Zool. xxv. 

 1896, pp. 413-418), shows that, besides the blood-supply, these receive also a tracheal 

 supply of peculiar character, and suggests that the fresh supply of oxygen is absorbed 

 from the water through the medium of the protoplasm of the hypodermis of the gill 

 in which the trachea? lie and passed directly to these trachea?, while the carbonic acid 

 is brought to the gill through the medium of the blood-flow and directly diffused to the 

 outer medium without any intermediate passage through the air-tubes. 



Iu the larva under consideration, the structure of the parts involved also points 

 to the conclusion that both the blood and the tracheal system play important parts in 

 the exchange of gases, and could Gilson's suggestion be proved to be correct, it would 

 throw a flood of light on this difficult question. 



The triangular processes (PI. 24. figs. 4 & 5) which enclose the area around the 

 spiracles bear along their free margins rows of stiff flattened hairs, those nearest 

 the apex being larger. When the processes are drawn together, an air-bubble is enclosed 

 and held by the bristle-like arrangement. The closing, however, does not invariably 

 occur when the animal is submerged. The fenestrated arrangement of the covers of the 

 spiracles will prevent the entrance of water into the air-tubes, while the blood-gills 

 probably serve for respiration when the animal is submerged. At the apex of each dorsal 

 process, and within the fringe, a specially long sensory hair takes its origin (PI. 24. 

 fig. 5, s.h.). It is articulated to its base of origin, so that its direction can be varied. 

 Their exact function was not decided. They appear to be sensitive to touch, but on 

 stimulating them the flaps were not invariably retracted. Perhaps they may serve to 

 indicate the necessity for folding the processes on submergence. 



Three other sensory hairs of different form also occur — one median, between the 

 above-mentioned one and the fringe, and two lateral ones, all within the area of the 

 fringe. They are somewhat shorter than the fringing-hairs and arise from a knob-like 

 articulation (PL 24. fig. 5 a). 



The arrangement of the chief dorsal branches of the tracheal system is shown in 

 PI. 27. fig. 19. 





