ELYMNIAS. By H. Fruhstorfer. 
385 
where Wallace met with it, and specimens from Perak are in my own collection. It somewhat resembles 
patna, and an approach to hypermnestra is noticeable on the under surface. There are two varieties of the 5; 
$-fa. mehida Here. 1. c. with white subapical bar on the forewings and white submarginal area on the hindwings, mehida. 
$-fa. abrisa, without these markings. — sumatrana Wall., described by its author from a 9 form, here figured abrisa. 
(88 c). $ differs from chelensis in the more uniform blue streaks on the forewings. Three variations of th q sumatrana. 
$$ must be placed on record; 1. $-fa. sumatrana Wall., with imposing white apical bar on the forewings, oc¬ 
cupying more than one fifth of the costa, thence passing towards the terminal margin, which it does not quite 
reach, so that a small part of the dove coloured ground remains at the extreme edge; in the cell between the 
two upper median nervules the ground colour sends an angular projection into the white area which is 
anteriorly exactly defined by the upper median. 2. £-fa. abrisa Dist., erroneously considered by its author 
to bead 1 , forewing unmarked, uniform dove colour, with an imperfectly defined, white, discal band on the hind- 
wings. 3. $-fa. immaculata Mart. Both wings with uniform dove coloured ground. — The underside in all immaculata. 
three forms harmonizes so completely with that of the. <§, that there can bo doubt as to their connection; but 
in the fa. sumatrana the white triangular marking on the costal margin of the forewing is most distinct and 
most like the while fa. abrisa has the least developed white spot on the costa of the hindwing. The brown 
tones described by english authors are entirely wanting in the Sumatra $$. From the apex of the forewing 
to the middle of the inner margin of the hindwing a shadowy line runs across both wings, having a certain re¬ 
semblance to the leaf marking on Kallima; the pale costal triangle on the forewings is placed inside this line, 
while on the hindwings the area beyond it is paler. Both sexes of the penanga form have indications of a se¬ 
ries of ocelli, indistinct, sometimes white centred, black dots; they are most distinct in the continental chelensis. 
— konga Sm. (= borneensis Sm., trepsichroides Shelf., penanga Btlr.). This is the most progressive race, konga. 
and the only one which can be recognized at a glance. The has 3 subapical blue streaks on the upperside 
of the forewing instead of 5, like the other local forms. There are 3 forms of the 1. the typical konga Sm. 
corresponds to the form abrisa Dist. from the Malay Peninsula and Sumatra. 2. mehiditia Fruhst. This $ mehidina. 
form Shelford describes as analogue of mehida, only the subapical bar on the forewings is narrower and more 
oblique. Expanse 65 mm. 3. fa. ptychandrina Fruhst. At present peculiar to Borneo, but it is not improbable ptychan- 
that similar $$ also occur in Malacca. Ptychandrina, apart from the absence of tails, is very similar to Pty- ina ' 
chandra schadenbergi Semp. from Mindanao in size, shape and markings. There is also in Staudingers collec¬ 
tion a $ analogous to mehidina, but with the cell entirely white in the forewings and the hindwings almost en¬ 
tirely white. North Borneo. 
With E. cumaea begins a series of four interesting species, striking from their size and beauty, as well 
as from their pronounced sexual dimorphism, and extremely difficult to separate, because both <$<$ an $$ 
possess many characters of the markings in common. All four occur exclusively in the Celebean subregion, 
and are separated into local races even in the main island. It is an interesting fact, that two species occuris, 
only in the north and east of Celebes, two only in South Celebes and its satellite islands, yet all four ap¬ 
pear in all their characters as coming from one source, whereby the celebean subregion of the great indo-mala- 
yan region shows itself unusually well defined. To this subregion belong, besides the main island, the satel¬ 
lite islands of Saleyer, Buton, Muna, Sula-Mangoli, Sula-Besi, Bankai, Peling, the Togian islands and the ab¬ 
solutely unknown Kambaena, which is most difficult of access, cumaea Fldr. (89 a) $ wings with peculiar, cumaea. 
light violet blue, slightly shining distal border. The $ is much larger than the <$, having an expanse of 49 mm 
as against 43—44 mm in the The ground colour is dark cacao brown with pale blue terminal border, 
which has a violet tint in the hindwings. The costal margin of the forewings is checkered white and pale 
grey. In addition there is a broad, whitish, violet tinged, oblique subapical bar, whereby the cumaea $ somew¬ 
hat recalls mimalon §, but the white dots found on the hindwings of mamalon and thyone are wanting. The 
under surface of all the wings is chiefly red-brown, with a distinctly paler subapical zone on the fore¬ 
wings, which is more extended than in the d'd*. The submarginal area of the underside of the hindwings 
is also lighter. Minahassa, North Celebes, rather rare, especially the $$. — toliana Fruhst., runs smaller toliana. 
than cumaea from Minahassa, distal border narrower, light grey-blue, rather darker than in bornemanni Ribbe. 
Under surface much lighter than in the typical cumaea, richly sprinkled with white, which is especially noti¬ 
ceable at the apex of the forewings. $ with the distal bordering more purple, in place of blue-green. Lower 
surface of $ leather brown than black. One $ has the lower surface strikingly light brown, and this colour 
also extends to the upper surface (pseudeuploea form, nov.), the dark brown is associated with whitish violet pseudeu- 
subapical bars on the forewings and yellow-brown subterminal area on the hindwings. Lower surface of hind- P^ oea - 
wings with 5 light blue shining spots besides the usual conspicuous white subcostal mark, which is so charac¬ 
teristic of all the four Celebes species. Toli-Toli, North Celebes, November, December; Tawaya, Central 
Celebes, August-September. — bornemanni Ribbe, is intermediate in shape and markings between hicetas bornemanni. 
and cumaea. The U bas on the upper surface of the wings just the same deep blue-black satiny colouring as 
the two species named. The bands on the terminal margin, which are blue, passing into grey in hicetas, 
are bright sky blue and considerably broader in bornemanni. In the fore wings the blue marking running along 
IX 
49 
