INTRODUCTION 
4 
For stability of nomonclature it is obviously necessary that the name given to the first described 
race of a species be adopted as the collective name of that species: but it should be clearly understood 
that the other races of the species are not derived from this firstnamed race but are of equal value 
with it: it is not the parent race, but only one of a number of allied geographical races. A local 
illustration should make this contention clear. The black and green Troides certainly originated hi 
New Guinea, and, spreading to the west, the south and the east, developed many geographical races. 
The first of these races to be named was 1\ priamns from Amboina: priamus therefore becomes 
the collective name of the species, though in itself of only equal value with eiiphorion from Cairns 
or the widely different croesus from Batchian. The race inhabiting Amboina is called 1\ priamus 
priumiis, and the race from Cairns 2\ priamus eiiphorion, and thus their relationship to a commou 
ancestor is at once declared. The four races developed in Australia are much too nearly alike to 
be considered distinct species, yet too divergent to be grouped under the one name: but T, priamus 
poseidoHj T. priamus pronoviuSt T. eiiphorion and T. priamus richmondius indicate clearly 
that they are all races of the one species, and also indicate their relationship with other races found 
beyond Australian limits. 
In insular races a specimen is occasionally to be found agi-eeing more closely with the nearest 
contiguous race than Avith that upon its own island: such examples are more frequent in a 
continental area where continuous land connection provides an easier means of intercommunication. 
Hence geographical races cannot be so limited as to include a particular form, without making 
allowance tor variation and toi‘ the inclusion of individual specimens resembling contiguous races: or 
to put it in other Avords, a geographical race cannot exclude any examples of the collective species 
found within its area. 
In order to form true opinions regarding the development of geograpliical races Avithin Australian 
limits, long series of accurately labelled specimens are absolutely essential: these our cabinets haA'c 
in most instances proA’ided. We have cai'efully studied our specimens in relation to their localities, 
and Avhere the great majority of a long series from one district shoAv definite points distinctive from 
those of other areas, Ave have considered that group of specimens to represent a race Avorthv of a 
subspeciiic name. But there are still some species of Avhich the knoAvn Australian examples are too 
fcAv to permit of a reliable opinion being formed: in some of these cases subspecies appear to 
be indicated, but Ave have not named them: Ave are opposed to the loolicy of founding new sub-species 
on scanty material, believing it preferable that a race should continue undetermined rather than the 
racial charactei's should be wrongly described, or a new race erected on insufficient grounds. 
In dealing Avith Avide-ranging species belonging to grouj^s that are not yet aa'gII kuoAvn, it is possible 
Ave may hav'e oA'erlooked some earlier name than that A\'e haA^e used for the collectwe species: Ave feel 
sure lioAveA’er that any such errors are feAV and far between, and AA^e are confident we have sufficiently 
indicated the Australian form to enable future monographers, Avorking out these little knoAVn groups, 
to correctly place the Australian race, AAffiene\'er Ave have not already done so. 
