390 
ELYMNIAS. By H. Fruhstorfer. 
brown. (Type from Waigiu in Coll. Fruhstorfer). The still nndescribed $ is in Staudingers collection and 
resembles a miniature Elymn. moranda (90 a). Ground colour black brown, forewings with greenish costal 
margin, hindwings distally slightly bordered with pale brown. Very rare, only 1 and 2 $$ known. 
viridescens. E. viridescens 8m. from Humboldt Bay is noted by its author as a closely allied species to papua Wall. 
kaJcarona. Hagen has discovered an allied race, kakarona Hay. on the Sattelberg in German New Guinea. forewings 
short roundish, expanse 32 mm. Upper surface dark velvet brown. Costa slightly glossed with steel blue. Fore¬ 
wings with a narrow bluish white streak, commencing on the anterior third of the costal margin and narro¬ 
wing inwards. Hindwings with light brown submarginal margin. Under surface dark brown, bar on forewings 
paler, broader; hindwings with a few transcellular, small blue spots. Forewings with a brown hair-pencil, 
springing from a yellowish pocket on the inner margin. Hindwings with a yellowish speculum, which also bears 
a brown hair tuft. 
If in ihryallis we find great female dimorphism, this condition is augmented in another, somewhat 
isolated race in the Papuan region, E. agondas, and its numerous subspecies, agondas has no trace of the fine 
striation, otherwise so typical of the whole genus, the contour of the wings is entirely rounded, and both d 
and $ have on the underside of the hindwings on the inner half of the terminal margin, a yellow band, in which 
are placed black, blue centred ocelli. The $$ are very light coloured, entirely pure white in the most extreme 
forms, excepting the margins of the wings, and have then a great resemblance to the Amathusiidae genus Tae- 
naris which inhabits the same region. The forms here included are among the most interesting and most pecu¬ 
liar butterflies of the Papuan Region. The indeed, do not differ essentially from their relations on the 
Moluccas (the vitellia-cybele group), the $$ on the other hand form a true paradigm for the most extended sexual 
dimorphism in their altered shape of wing and their inclination towards the Taenaridae. agondas inclines to 
the formation of local races, and joins on the the Taenaridae in this detail also. In general the races of agondas 
in New Guinea and the subjacent Islands resemble Taenaris bioculata Guer., with blue ocelli on the upper sur¬ 
face of the hindwings; only melagondas Fruhst. is an exception, its marking and the distribution of the colours 
recalling Taenaris mailua Sin. from the same locality (Convergence). All agondas have a strongly shining friction- 
spot on the under surface of the fore wing, as well as a yellowish, cpiadrate band at the base of the forewing, 
similar to that in mimalon Hew. The corresponding speculum on the costal margin of the upper surface of the 
hindwing is partly covered with red-brown, slightly shining scales. On the upper border of the cell are two 
tufts of long hairs, which cover a large, elongate, androconiapelt. The colour and mode of arrangement of the 
androconia, as well as the colour of the scent tufts, differ slightly in the individual local races; these scent tufts 
are usually composed of entirely black hairs, only in melagondas the hairs of the proximal tuft are basally red- 
agondas. brown, apically whitish. The distribution of the agondas races is as follows; agendas Bsd. Boisduval figures 
an unicolorous dark green specimen and gives Vanikoro, the southermost of the Santa Cruz Islands, as the 
habitat of his type, but agondas has not been found recently more eastward than Woodlark, so that we must 
not expect to meet with it in Micronesia, Polynesia or the Solomons. It is much more likely than agondas ori¬ 
ginates from a papuan island near Waigiu, or even more probably, from the Dutch part of the main Island 
itself. Compared with my material, the figure harmonizes best with specimens from Sorrong, the north-westerly 
corner of New Guinea. H. Kuhn has collected there repatedly, and rediscovered several forgotten old types, 
as for instance, Taenaris artemis, Voll. and T. dioptrica, Voll. We may therefore, for the present transfer the 
still doubtful locality to Dutch New Guinea. $$ from Sorrong have a narrow, black apical margin to the fore¬ 
wings, which is also slightly brownish as is the distal area. Hindwings with tAvo large oblong, only narrowly black 
ringed blue eye spots, a black brown, comparatively narrow distal margin, which is proximally sliglAly lighter 
brown. The similarity of the agondas $$ from Sorrong to Taenaris bioculata pallida Fruhst. from the same 
locality is fascinating. Only a few degree of latitude eastward, at the foot of the Arfak mountains in Geelvink 
bioculatus. Bay we meet with a new $ form. This answers best to bioculatus Dbl.-Hew. (89d 1, misprinted biocellatus). The 
broad, black-grey frosted outer area of all wings somewhat resembles the same appearance in Taenaris biocu¬ 
latus charonides, Stgr. from the far more easterly Humboldt Bay. I have unfortunately no specimens of bio¬ 
culatus from Dorey, but I am convinced they would have the same distribution of colouring as bioculatus. — 
In German New Guinea we meet with a geographical offshoot so greatly differentiated, that Staudinger took 
glaucopis. it for a separate species, and described it as such; glaucopis Stgr. (= melanippe, Sm.) a migrant from the South. 
rare on the Sattelberg and Finschhafen, not occurring in Stefansort and Friedrich-Wilhelmshafen. The $$ 
of this form do not follow their Taenaris model so closely, their almost entirely black hindwings being more 
like bioculatus charon Stgr. from British New Guinea and on the hindwings Taenaris dina Stgr. from German 
New Guinea. The of glaucopis vary from the Ud from Dutch New Guinea in the presence of a narrow orange 
subanal band on the hindwings, which are more elongate and accompanied by a brightning of the forewings 
melagondas. in melagondas Fruhst. (89 d $, 90 a U)- Specimens from Collingwood Bay are rather darker than those from 
Milne Bay, where examples with greenish white entire submarginal band on the forewings are not uncommon. 
The of melagondas recall Taenaris mailua, Sin. on both upper and lower surfaces. CollingAA'ood Bay, Milne 
