IN THE LARViE OF ANISOPTERID DEAGONFLIES. 



169 



o£ respiration. In this respect the Undulate Type is distinctly at the 

 greatest disadvantage, the Lamellate Type, on the other hand, easily the 

 most advantageous. Forms in which papillse are developed clearly have an 

 advantage over similar forms without papillse; but it is doubtful whether a 

 single folia of Anax, with all its papillse, is of any greater value in extracting 

 oxygen than is a single gill of the Lamellate Type. 



The following table gives a comparative summary of the number of 

 capillar}^ loops calculated for the different genera examined. The totals are 

 only approximate, but the degree of error in each is probably very much of 

 the same order : — 



Origin of the Lamellate Type. — It is a matter of great regret to me that, 

 during the year in which this research has been carried out, I have been 

 unable to obtain any Libellulid ova for hatching out the young larvae. The 

 smallest larvse that I have been able to obtain were of the genus Diplacodes, 

 and measured from 2 to 3 mm. in total length of body. I consider these 

 larvse to belono- to the third or fourth instar. Several of these were sectioned. 

 The results showed, not only that the Lamellate Type of gill was already fully 

 established, but that basal pads were already fairly well formed. In no case 

 was I able to discover any sign of the main longitudinal gill-folds of the 

 Simplex System, which are the principal breathing-organs in the young 

 larvse of urEschia and Anax (p. 143). It is, therefore, clear that the Lamellate 

 Type, though necessarily classed within the Duplex System because of its 

 structure, has no close phyletic connection with those other types of the 



* Minimum. 



t Maximum. 



