140 
MEMOIRS OF THE QUEENSLAND MUSEUM. 
AUSTRALIAN 
HYMENOPTERA CHALCIDOIDEA-IV.* 
The Family Eulophidae with Descriptions of New Genera and Species. 
By A. A. Girault. 
INTRODUCTION. 
* 
Of all the families of vvliat is undoubtedly the largest group numerically of the great 
order Hymenoptora, this, the Eulopiiidat, seems to be the representative one in the Australian 
fauna, the number of genet a and species exceeding these of any other family of the great 
complex. This paper records but lew of the species but perhaps most of the genera existing 
to-day on the Australian continent. The groups or tribes established by Ashmead (1904) 
have been adhered to because they seem to be natural but a large number of new genera have 
been established more especially in the Tetrastich into. These genera I believe to be natural 
groups but for the present they are established especially because it seems necessary to do so 
in order to avoid that greatest of difficulties with these minute insects, namely, identification 
*of specific forms. It has been true in the past that identification was thought to be insured 
by a general description of the species, omitting the generic characters since these were 
understood as being present. Rut this practice has failed signally in establishing identity of 
species, for the reasons that some species are carelessly or otherwise referred wrongfully to 
genera, others differ somewhat from the generic description and these differences are not 
pointed out while still others differ in smaller characters, those not included within the descrip- 
tion of the old genera and not being brought to the attention of the student are, at the time, 
overlooked. The need and use of revisions show how common and ordinary are the great 
number of omissions made in description# of species and how helpless the student becomes 
when attempting identification upon these incompdete diagnoses. All of these serious difficul- 
ties will be obviated by this what is called (wrongfully, I think and hope) splitting. The 
student is forced to be careful, the generic diagnosis is as concise as nature allows, specific 
characters are readily grasped and any variations from the genotype easily and quickly 
noted. These are the conveniences afforded by the establishment of these genera. Identifica- 
tion is the basis of nomenclature. 
On the other hand, nomenclature is the tool or convenience of classification which is 
concerned with heredity. The tribes of the Eulophida) seem natural but it is difficult to define 
natural geuera; yet going upon the principle that genera arc to species as species are to 
varieties and so on (that is genera are groups comprised of clusters of species, groups within 
a group) we may say that a genus is natural and valid or good when all of Its species or 
groups of species are mutually inclusive and do not vary enough from each other to intergrade 
with the species of other mutually inclusive groups (or other valid genera). Thus, Tetrastichus 
Haliday and Aprostoceius Westwood are held to be valid genera differing in but one jiarticular, 
the number of the ring-joints of the antenna*. E’ut if there should be found a species of 
either group in which the number of ring-joints Avas inconstant, intergrading is established 
and the two geuera must, be considered as one, since their inherited structure is the same. 
This is the principle upon Which I have based these genera, in other language they are based 
upon the constancy (or supposed constancy) of a character or of characters which in turn are 
♦Contribution No. 15, Entomological Laboratory, Bureau of Sugar Experiment Stations, Bundaberg, 
Queensland. 
