180 
PAKEBONIA. By H. Fruhstorfer. 
bocbera. 
phazania. 
hemara. 
arscnnota. 
bazilana. 
lilacina. 
persidina. 
elaitia. 
tritaea. 
caecinia. 
bargylia. 
hcrmocinia. 
bilinearis. 
octaviae. 
already mentioned, is somewhat doubtful, but the genitalia show rather important differences. The uncus 
of pingasa is straighter and distally less curved than that of the Valeria series of forms, the valve much 
broader, more rounded and distally less pointed. 
P. boebera. A species of somewhat doubtful validity, parallel to pingasa in the broad scent-patch 
of the hindwing, but on the other hand in the very variable $$ so closely allied to Valeria that it may 
perhaps be regarded as only its most northerly branch. But as the forms of boebera form a natural group, 
they may here be provisionally treated as belonging to a collective species. Like the continental hippia 
and persides boebera varies on the Philippines both according to the locality and the season, the latter, on 
account of the greater climatic variations, more on the northern than on the southern islands. All the $$ 
of the dry cold season are entirely white-spotted, those of the rainy season somewhat yellow with more 
extended black markings. Transitional forms have a bluish white sheen. As we go from north to south 
the black distal border of the <$<$ increases in width, so that the boeberci-SE of the South Philippines 
approximate to those of phocaea (66 e) and according to Semper it is even difficult to separate the $$ 
of the two species, boebera Eschsch., of which we figure the <3 and $ of the rainy-season form (phazania 
form, nov.) (66 c), has both sexes of the dry-season form essentially smaller, with larger white subapical 
spots on both wings, the lighter blue, the $ dull green-white. Luzon. -— hemara form, nov., found in 
North-West Luzon at Yigan, is an intermediate form, Avhich Semper figured (Schmett. Philipp, pi. 42 
fig. 7) and which differs from the cold-season brood in having the bands narrower and yellowish instead 
of whitish. — arsamota subsp. nov. is smaller than even the winter form of boebera and is beneath only 
slightly suffused with blue-grey with very long but narrow transcellular stripes on the hindwing. Island 
of Negros. — bazilana Fruhst. approaches pingasa Moore from South India in the broadened black margins 
of both wings. The are suffused with dark green instead of light blue, the forewing bears either no 
spots at all or only one subapical white patch. The $$ beneath blue-black almost to the middle of the 
wing. Three forms of the $ are united in the Fruhstorfer collection, that of the rainy season 
together with a transitional form with blue-violet sheen on both wings (f. lilacina form, nov.) and an 
apparently extreme dry-season form, recalling hemara from Luzon and persides-Q from Annam, with 
reduced white-grey transverse stripes on both wings, which are densely dusted over with black (f. persidina 
nov.). Bazilan, near Mindanao. — elaitia subsp. nov., the darkest of all the known races, distal border 
of the hindwing extended to the middle cell, subapical spots of the forewing entirely absent. Island of 
Panaon. 
P. tritaea is the largest species; its origin is perhaps to be traced to Mindanao, at the head of 
the ancient bridge of land connecting the Philippines with Celebes, unless conversely it was rather its 
representative phocaea which came from Celebes to the zoologically allied Mindanao. On Celebes itself the 
species is split up into two essentially modified subspecies, of which tritaea Fldr. (67 a) is characteristic 
of the edges of the woods in Northern Celebes. In Central Celebes and in Gorontalo at the eastern arm 
of the island flies in the months of August and September a smaller dry-season form with somewhat purer 
and larger white streaks and spots and lighter grey-brown tinge on the under surface of the $$ (f. cae¬ 
cinia form. nov.). In Southern Celebes not only the pattern is different but also the shape, especially the 
forewing being more rounded, in the $$ the brown-violet tinge on the upperside of both wings is absent, 
and the submarginal spots are almost twice as broad, being particularly prominent on the underside of the 
hindwing in the shape of beautiful, white star-shaped patches. We name the South Celebes race bargylia 
subsp. nov. Common at the waterfall of Maros. — On Bangkai occurs a similar, still smaller race, hermo- 
cinia subsp. nov., with darkened stripes in the cell of the forewing, reduced submarginal dots and obsolete 
circumcellular spots on the hindwing. — bilinearis subsp. nov., in size a transition from tritaea to the next 
island form, differs from both in having the black stripes of the cell prolonged to the base of the fore¬ 
wing, being hence almost as long again as in tritaea, whilst in octaviae they are absent. Hindwing almost 
without basal, i. e. much abbreviated blue-green longitudinal stripes inside the cell. Salaycr, discovered by 
Dr. Martin. — octaviae Snell. (67 b) has ES with or without short black stripes before the apex of the 
cell of the forewing. The ground-colour lighter blue than in tritaea, $ approaching that of bargylia Fruhst. 
from South Celebes, with lighter transverse bands and shorter black lines in the cell of the hindwing. 
Under surface of the <$$ without a trace of black tinge, brilliant blue-white and with a strong mother-of- 
pearl gloss. Subapical stripes of the $$ on the underside of the fore wing essentially broader than in the 
Celebes tritaea- races, hindwing margined with lighter brown. Common on Tana-Djampea and Kalao, 
especially in December. 
P. phocaea Fldr. from the South Philippine region must perhaps be regarded as a subspecies of 
the gigantic tritaea Fldr. from Celebes, a question, however, which can only be decided with the help of 
larger series, especially of $$. The genitalia, for instance, are somewhat different, the uncus of phocaea 
decidedly shorter than that of tritaea Fldr. phocaea is very constant according to Semper, who Avas able 
to examine almost 900 specimens. The $$ always bear in the cell of the hindwing above a black longitu¬ 
dinal streak, which is but rarely present in the boebera and Valeria races. The strongly convex costal 
