154 



tinguished is very great as compared with those so knowTi in 

 1895. For this advance we are principally indebted to Needham 

 (1901-1903, etc.) for North America, Lucas (1900) in England, 

 Rousseau (1908, 1909) in Belgium, Ris (1909) for Switzerland 

 and German3% and Tillyard in Australia. These specific 

 identifications were chiefl}' made by actual rearings of larvae to 

 the imago. By microscopic examination of the rudimentary 

 larval wings, Needham (1904, etc.) has been able to recognise 

 the venational peculiarities of the future adults, and to determine 

 the species by this method, which has been employed by other 

 students also. 



Fossil Odonata. 



Meunier has photographically figured (1896-1898) a number 

 of specimens from European museums, and Cockerell (1907,. 

 1908) has made known some interesting forms from Florissant,. 

 Colorado ; but for the chief advance in this field, we are indebted 

 to the voluminous handbook of Handlirsch, Die Fossilen 

 Insekten (1906-1908), of wide-reaching view, from which we 

 have already quoted in dealing with the classification of the 

 Odonata. Handlirsch enumerates nine species of Paleozoic 

 Protodonata from the Upper Carboniferous and the Permian of 

 Europe (no Odonata being known from this series of rocks, nor 

 are the Protodonata known from any later epoch), and of the 

 Odonata proper, sixt^'-seven Mesozoic and ninet^'-two Tertiary 

 species. 



The Protodonata are considered to be ancestral to the Odonata 

 and derivable from the still more ancient Palaeodictyoptera, from 

 which latter they difíer by the narrower inter-alar tergites, the 

 fusion of the basal parts of some of the longitudinal veins of the 

 wings, the transformation of many longitudinal veins into so- 

 called interposed sectors which apparently arise from cross- 

 veins, and the presence of numerous, more regularly arranged,, 

 straight cross-veins. 



On the other hand, the Protodonata difíer from the Odonata 

 by the lack of nodus, of pterostigma, and of the crossing of the 

 radial sector over the first two branches of the media.' 



1 This last difference is denied by Sellaros (1906). 



