150 



to 22 mm. He was unable to correlate the size, the number of 

 moults, the length of larval life, and the temperatures to which 

 the insects had been exposed, except to point out that there is 

 a tendency " for the larger nymphs in any stage to be those 

 which ultimately complete in the smallest number of stages." 

 Ischnura elegans passed through twelve larval stages ; other 

 species of Agrionines were thought to have from eleven to four- 

 teen. Much information is given on the rate of growth of body- 

 length, of the antennie, the labium, and the wing-rudiments. 



Backhoff (1910), in the paper airead}^ quoted, estimated the 

 number of larval stages in Pyrrhosoma nymphîda {minium) at 

 nine, from comparisons of the sizes of different larvae. East 

 (1900) observed the last seven moults in the larva of Mshyia 

 cyanea and inferred a total of nine, ten, or eleven. E. M. Walker 

 {1912) observed the last eight moults in Canadian Mshnœ and 

 believes they are preceded by three or four others, making a 

 probable total of twelve or thirteen larval stages, occupying 

 probably three years. 



Certain Larval Limitations and Peculiarities. 



R. C. OsBURN (1906) conducted a series of experiments to 

 determine the ability of eggs and larvae of various species of 

 Odonata to live in sahne solutions of diñerent densities, but 

 none were able to withstand a density over i"oi. He demon- 

 strated that there exists " a ver}^ definite barrier to their assump- 

 tion of marine life, and that this barrier remains unchanged 

 during the life of the individual." 



Perkins (see Sharp, 1895) (1899), in the Hawaiian Islands, 

 and Knab and Calvert (1910, 1911), in Mexico and Costa Rica, 

 have found that certain Agrionine larvae [Agrión and Mecis- 

 togaster respectively) live in the small accumulations of water 

 between leaf-bases of terrestrial and epiphytic plants, or, in the 

 Hawaiian case, even between dry leaves. Similar habits on the 

 part of vmdetermined Agrionid and Libellulid larvae in Malaya 

 have also been reported by Leicester (1903). The food of such 

 larvae consists of the early stages of mosquitoes, of other Diptera, 

 etc., hving in the same situations. 



The discovery of Petalurid larvae {Tachopieryx by William- 



