104 CAUDAL GILLS OF ZYGOPTEfilD LARV.S, 



session of an internal structure which fits them to act more or 

 less etfectively as respiratory organs. They all possess a well- 

 developed ti-aclieal system, tM o large blo(xl-canals, and a definite 

 accompanying nervous-system. Of these, the most important, 

 from the point of view of respiration, is undoubtedly the tracheal 

 system. I propose, therefore, to restrict the term Reduced (Non- 

 functional) Type to those gills in which, of whate% er form they 

 may be otherwise, the tracheal system is so reduced that we may 

 reasonably claim that they no longer function as tracheal gills. 



The only gills of this type which have come under my notice 

 are the extraordinary gills of the plant-dwelling larva of Ayrion 

 astelue Perkins, from Hawaii. AVe have already seen (p. 98) that, 

 in the genus Agrion, we find certain larvae with well-developed, 

 vertical, lamellar gills of the nodate form, while others have their 

 gills of denodate form. The members of the genus Agrion which 

 have colonised the Hawaiian Islands all appear to have adopted a 



Text-fig. 30. 



Reduced (non-functional.) gills of Agriwi astd'nv Perkins, in natural posi- 

 tion, lateral view; ( x 18). 



new habit of larval life. Their larvje. no longer dwell in permanent 

 waters, sucli as rivers and pools, but pass their lives hiding at 

 the bases of the leaves of various forest-plants, where water from 

 tlie hea\y rains collects, and where, apparently, a plentiful 

 supply of smaller insect-larvte is also to be found for food. The 

 best known of these larvae is that of Agrion asteJkf, of which I 

 have fortunately been able to examine a large number, sent to 

 me bv Mr. J. C. Bridwell, of Honolulu. 



