151 



SON, 1901 ; Petalura by Till yard 1909), and of that of Petalia 

 (Till YARD, 1910), have been noteworthy events. Those of the 

 former group hve in boggy swamps, and in the case of the 

 Austrahan Petalura in foul, mudd}- ooze which cakes sohdl\- on 

 the larva at the time of transformation. A mud-dwelling habit 

 is also that of the larva of the Australian Synthemis eustalacta, 

 which, encased in dry mud, has been known to resist drought for 

 ten weeks, and to fast for three months (Till yard, 1910). 



Taxonomy. 



A comparative study of Odonate larvic and their corre- 

 sponding imagos led Ris (1896) to what maybe termed the first 

 phylogenetic classification of Odonata based on the ontogen\' of 

 an organ followed throughout the entire order. This organ was 

 the gizzard. Ris showed that the larvie of the Calopterygina? 

 and some Agrioninae have the most complexly armed gizzard, the 

 teeth being arranged in sixteen longitudinal folds, reduced to 

 eight in Lestes, to four in Gomphus and /Eschna, and further 

 reduced to four bilaterally symmetrical teeth (not folds) in 

 Cordulegaster and the Libellulidai. He also showed that a reduc- 

 tion in the folds and in the teeth takes place in the develop- 

 ment of the individual in each of these groups, scarcely more 

 than traces of the teeth remaining in the imagos of the Anisop- 

 tera. Combining these results with data drawn from other 

 organs, some of which had already been employed by previous 

 writers, Ris presented his ideas of the relationships of the sub- 

 families of the Odonata in the form of a genealogical tree, con- 

 fessedly a modification of that of Calvert (1893), from which it 

 difíers chiefly in placing the Cordulegastrinai as ancestral forms 

 of the Libellulidfe. 



Both Calvert's and Ris's views agreed in regarding the 

 Zygoptera as more primitive^ and in looking on the Caloptery- 

 ginie as ancestral to all other Odonata. Needham, and his 

 students Miss Butler and Thompson, whose work has been 

 quoted above, have made many suggestions as to the generalised 

 or specialised conditions to be found in various structures ol 

 both the Zygoptera and Anisoptera, but have refrained from 

 formulating any detailed statemmt of the relation>hijis of thesi' 



