TACHIN1DAE OF AUSTRALIA 27 



of the subfamilies that should be recognized and on the combinations of characters 

 that should be used to define them. Most taxonomists currently recognize relatively 

 few subfamilies (normally between four and six) but a large number of tribes, but 

 none of these family-group taxa are susceptible of simple and unequivocal definition 

 that will render their members immediately identifiable on straightforward key 

 characters. The existence of many aberrant forms, and of much convergent 

 evolution, compounds the classificatory difficulties still further and ensures that - 

 view them how we may - it is simply impossible to define the subfamilies succinctly 

 or to key them out (at least on the external morphology of the adults) with any 

 precision. Undoubtedly, however, recent and continuing studies on the male 

 genital structure, the female reproductive habit, the morphology of the larvae 

 (especially the first instar larvae), and of the host relations, are doing much to 

 enlighten specialists on the probable phyletic relationships, and hopefully this newly 

 acquired data may lead to more concrete subfamiliar definition in the course of time. 

 Meanwhile, it is possible to recognize many of the members of the more distinctive 

 subfamilies (Phasiinae, Goniinae) on their overall facies, even though it is extremely 

 difficult to fix their characters in an unexceptionable way that is satisfactory for 

 key construction, and a knowledge of the host relations of any tachinid parasite 

 can be of immense value in placing its subfamiliar identity - for example, any 

 Australian tachinid with a host in the Hemiptera belongs in the Phasiinae. 



In the present work four subfamilies are recognized, the Phasiinae, Proseninae 

 (=Dexiinae), Tachininae and Goniinae, but it is freely admitted that the Tachininae 

 in the sense here used is a heterogeneous assemblage of forms in which two or three 

 distinct subfamilies ought probably to be acknowledged; but until a clearer picture 

 of the interrelationships of the included forms emerges it remains useful to treat all 

 the Australian tachinids that are clearly not either Phasiinae or Proseninae or 

 Goniinae as constituting one subfamily (for which the name Tachininae is nomen- 

 claturally correct). 



With the difficulties in defining the subfamilies, and the confusing overlap in 

 many of their characters, it is impossible to construct a workable key even to the 

 small number of subfamilies here recognized that will permit each and every specimen 

 to be placed unequivocally in a subfamily. The key that follows is only a tentative 

 guide to the probable suprageneric grouping to which any specimen belongs. In 

 order to keep the key short and simplified (without long confusing couplets that 

 attempt to cover every exceptional or aberrant form) certain tribes have been run 

 out individually and sometimes there is more than one exit for a particular sub- 

 family. Some extremely poorly known forms of very uncertain subfamilial position 

 (e.g. Myiotrixa, Amphitropesa, Neximyia) have been omitted from consider- 

 ation, as to include them would so complicate the key as to negate much of its 

 value. When the names of taxa are juxtaposed this does not imply close phyletic 

 relationship. 



1 Ocelli absent. Prosternal region grossly inflated, visible from side view. [Probably 



parasites of Orthoptera] ...... Ormiini (Tachininae, part) (p. 57) 



- Ocelli present. Prosternal region normal, not visible in profile .... 2 



