388 
ELYMNIAS. By H. Fruhstoreer. 
whiter subapical bar on the forewings, which continues along the terminal margin as a series of spots, sumptu- 
atys. osa is thus probably the most northerly form of hewitsoni. — atys Fruhst. The $ approaches sumptuosa <$, but 
the terminal margin of the wings violet instead of white. Ocelli on hindwings not broadened. Ground 
colour more violet, not black or greenish blue as in hewitsoni. submarginal band on forewings paler, more 
conspicuous. Ocelli on hindwings more prominent. Under surface lighter brown, more richly white marked. 
mrliophila. Wings rounder, habitus smaller than in hewitsoni. East Celebes, Tombugu and vicinity of Bonthain. — me- 
liophila Fruhst. (90 a). A satellite-island race of such greatly altered appearance, that I first took it to be 
quite a different species, and placed it with cumaea Fldr. until recently the four, certainly only rudimentary, 
white submarginal spots on the lower surface of the forewings made me doubtful and set me on the right track. 
It seemed also suspicious, that cumaea should have a form on Saleyer, whereas it had not yet been observed 
in South or East Celebes. Runs smaller than hewitsoni, upper surface dark steel blue with violet tinge. Hind¬ 
wings, upperside with 4 whitish violet spots, forewings with slight, pale grey green apical tint, otherwise unmar¬ 
ked. Under surface rather duller black-grey than hewitsoni, less white marked, instead of the large white ocelli 
only 4 inconsiderable spots on the forewings and 7 similar ones on the hindwings, the A is less decorated and 
more uniform coloured than hewitsoni $ and has only four small ocelli on the hindwings. The chief difference 
is in the $, in which the terminal margin of the forewings is fiery yellow-brown, with blue grey iridescence with 
which the three greenish white subapical spots are finely contrasted, whereas the terminal margin of the hind¬ 
wings, which has five white ocelli, is tinged with violet, a truly grand butterfly. Salayer Island, taken by 
me in March. 
To the east of Celebes the species of Elymnias become less numerous, only two species usually, some¬ 
times but one on some islands. The contour of both wings, already considerably rounded in the Elym¬ 
nias of Celebes, becomes rounder and more even in those of the Moluccas and Papua, the former of which retain 
the peculiar streaking only on the under surface of both wings along the terminal margin, while the latter have 
no trace of this marking which is so typical for the whole; genus. The following sexual characters are common 
to the Celebes and Molucca species; forewing, a brand of pale yellow scales below the base of the submedian, 
which is distally much curved. A very large, strongly shining and depressed sex mark from the inner mar¬ 
gin to the cell. Upper surface of hindwing; a distinct precostal cell, similar to that in the Brassolidae. Rather 
below the origin of the radials are two tufts of stiff black hairs, which cover two oblong patches of scales covered 
with thick black androconias. The covered part on the costal margin palish brown. In holofernes Btlr. and 
cumaea phrikonis Fruhst. the hair tufts are shorter. The South and North Moluccas are each inhabited by 
one species, which can be easily separated from one another. 
In E. cybele the North Molucca species, the white ocellar submarginal spots are placed exactly parallel 
with the terminal margin of both wings, whereas in vitellia Cr. the South Molucca species, they have a superficial re¬ 
semblance to a constellation, which is produced by their irregular, distorted position. The fresher the speci¬ 
men, the darker the ground colour of the under surface, and the more distinct the appearance of stars; Butler 
named the form “ astrifera " having undoubtedly received the same impression. On the forewings the three 
central of the five ocelli are displaced towards the base, so far that the whole series forms an inwardly angled 
line; the fifth lowest spot stands at the same distance from the margin as the first, uppermost. As an indica¬ 
tion of relationship the otherwise rather straight series in cybele has also a slight inward convexity. The hind¬ 
wing with its 6 ocelli is similarly conditioned; in cybele a regular curve, parallel to the margin of the wing, in 
vitellia a straight line from the upper costal to the 2nd outermost apical spot, from which each further spot 
recedes more inwardly in a rather straight row, the 4 and 5 to the greatest extent, so that the line shows a 
slight inward convexity. If we take the white spot at the apex of the cell as a starting point for the whole 
series, the spots take the form of the upper half of a note interrogation or of a 2. For the rest, the contour of both 
cybele. wings in cybele is one degree rounder than in the still scalloped vitellia. — cybele Fldr. from Batjan, especially 
astrifera. common in August, but 9$ very rare. Butler gave the name of astrifera to small specimens, which may re- 
opaca. present a dry season form. — opaca Fruhst. larger than the last named but smaller than cybele. White mar¬ 
king of the under surface less conspicuous. $ darker and uniform brown, without the reddish tinge on the me¬ 
dian area of the upper side of the wings. The white submarginal spots on the under surface much more deli- 
ternatana. cate than in cybele. Halmaheira. — ternatana Fruhst. Smaller than cybele and opaca. Ground colour 
uniform light redbrown, which is also noticeable beneath. White marking and striation of the underside of 
obiana. the hindwings much reduced. Ternate, May. — obiana Fruhst. (88 e) departs further from the name type than 
any other insular races of the North Moluccas, the 9 especially presents a splendid appearance with the blue- 
violet subapical marks on the forewings, reminding one of the of E. vitellia Cr. (88 d). obiana is also easily 
recognizable on the lower surface by the large white submarginal stars; the distal portion of the under surface 
on the fore and hindwings is not black, but light brown, and the 9 has in the cliscal part of the upperside of the 
forewings a broad, light brown spot, which passes into the dark brown ground colour beyond the subme¬ 
dian. By means of this spot obiama somewhat recalls thryllis Kirsch $ from New Guinea, and certain Euploeas 
(Sarobia confusa, Btlr.). cybele obiana has nothing in common with vitellia Cr. since in obiana the white sub- 
