204 



ing the value of Casteliiau's obsisrvations on tliese insects, I can 

 only plead that I thought M. Putzeys' evidence that all the 

 species of Farroa in the Castlenau collection when he went over 

 them in 1873 were females, was conclusive proof Castelnau 

 was in error in attributing to the males of Parroa nonsecuriform 

 palpi and non-spongiose anterior tarsi. 



I have now seen three species in which the male is certainly 

 without spongiose tissue on the underside of any joints of the 

 tarsi ; so it is evident Castelnau was right. Either such male 

 specimens must have been lost before M. Putzeys looked through 

 the Castlenau collection ; or he spoke after a mere superficial 

 examination of the specimens before him, when he said all the 

 species of Parroa were females. 



I am now able to make a considerably more extended review 

 of the forms comprised in the genera Parroa, Adotela, and 

 Cerotalis than in 1890; but my knowledge of the previously 

 described species is confijied to so few species that T can suggest 

 no alteration in their classification, nor can I find any decided 

 differences, apart from sexual characters, between them. Genera 

 founded entirely on features appertaining to one sex seem to me 

 too artificial to be maintained in a natural system of classifica- 

 tion ; but, when our knowledge of a group is too slight to enable 

 us to divide it into genera of undoubted natural value, an arti- 

 ficial system must be adopted, and this seems such a case. As 

 will be seen from my remarks on the genus Parroa, which follow, 

 I am doubtful of the position of many species ; but the matei-ial 

 I have clearly indicates that, either all the species now placed in the 

 genera Parroa, Adotela, and Cerotalis will have to be united- in one 

 genus, or new genera will in all likelihood be required for Parroa 

 (Adotela) aitstralis, SI, and similar forms; for Parroa apicalis, SI.; 

 and possibly for Adotela Frenchi, SI. There seems little doubt 

 but that we may expect many new species of Broscini to be 

 found in Australia, which, when obtained, will enable us to deter- 

 mine the true classificatory value of the differences between 

 forms such as those on which tlie genera Parroa, Adotela, and 

 Cerotalis have been establislied. In the meantime, to make new 

 genera for apparently aberrant forms would in all probability 

 merely cause confusion and instability of nomenclature, troubles 

 already oidy too rife in zoology. 



Genus Parroa. 



I regard P. Hovntti, Casteln., as the type of the genus 

 Parroa ; with it may be associated P. grandis, Casteln., and P. 

 noctis, SI. (n. sp., vide post). These species seem united by a 

 similarity in the form of the labrum, antenniB, femora, posterior 

 trochanters, tarsi, &c.; but, as the male of neither P. yrandis nor 



