CASTALIUS. By H. Fruhstorfer. 
883 
character different from the next allies. The neuration differs from one species to another and resembles some¬ 
what that of the Nacaduba-group, without, however, crossing the first subcostal as in them and in the Everinae. 
In general the first subcostal closely approaches the costal and, according to the species it is more or less coales- 
cent with it, whilst in the Tarucus it is separate, though without being just as distantly away from it there as 
in Taraka or the genuine Lycaena. Thus we must quite at libitum insert the Castalius there where de Niceville 
had already placed them in 1890, between Catockrysops and Polyommatus. 
The morphology of the sexual organs, however, offers us still less support than the neuration, and in 
Asia there are probably only few genera that are anatomically jumbled up in a similar way as just the Castalius 
are. We find in them the most primitive organs ( Taraka, Upolampes) and by their side the most grotesque 
formations to be thought of in the Indo-Australians {Cast, fasciatus, Callictita cyara), if we take in view the 
armature of the valves. Nor does the shape of the oedeagus offer us any clue to allied groups of species, although 
three species {C. elna, roxus, caleta ) possess such an uncommonly clumsy, mortar-shaped oedeagus as it does not 
exist in any other group of the Lycaeninae. But this apparent' differentiation being applicable systematically 
or generically is again lessened by the less monstrous formations in C. mindarus, C. ethion. Two more, quite 
pure, genuine Castaliinae, however, have such a normal slender penis that by its construction it is most closely 
allied to the genus Lycaena and presumably represents an organ of the lowest grade of development, if we do 
not want to consider it as an atavism to the great original type. — Thus, the more species we take into consi¬ 
deration, the more the generic characters become effaced. Nearly all our genera are crutches by means of 
which we only slowly make headway. Nevertheless the anatomy has even supported some genera ( Taraka , 
Upolampes) and cancelled but one superfluous genus ( Cyclyrius ), as it has also otherwise cleared up the close 
affinity of the species. The appliance of the anatomic method again led to a complete success also in the Castaliinae 
since by its aid was comprised what remained separated by the former authors (races of the Cast, caleta- group), 
or separated there where too much was united. Apophysis lateralis (Reverdin) or Falces (Beth. Baker), in 
contrast with the genera Nacaduba and Lycaenopsis, are a very common appearance in the Castaliinae and 
they exist more or less completely in nearly 50 percent of the species. It was, however, impossible to group 
the species according to the presence or absence of these secondary appendices, because species otherwise closely 
allied would have been torn asunder, entirely distant species, however, united. Androconia occur only in one 
species from New Guinea, as also among all the African Castaliinae male marks were only found in the magnificent 
C. isis Drury. — If we consider all the results of the comparative morphology in their relations, we find that 
the Castaliinae, in contrast with the almost unequalled uniformity and regularity of the markings on the under 
surface of the wings, represent, in an anatomical respect, an entirely heterogeneous mixture. But thereby 
the Castaliinae deviate from the Thy-sonotis, Lampides, Poritia exhibiting many features common among each 
other. Also in their geographical distribution we notice the same irregularity. We find among them wide¬ 
spread ubiquists {C. rosimon) beside entirely scattered, isolated elements (C . fasciatus). On the whole the Casta¬ 
liinae must be regarded as inhabitants of the lowlands, although some species are known to be met with yet 
at altitudes of about 1500 m. Except the Papuan Callictita, not one species inhabits exclusively the mountains. 
Apart from Africa, we find the Castaliinae distributed from India, the Andamans and Nicobars to New Guinea. 
In the Salomons the genus is absent, even the Bismarck Archipelago is reached by only one species, whereas 
of the allied Tarucus one species penetrates to the Pacific Islands. Most of the species are entirely tropical, 
only the subgenus Taraka passes over also to the temperate zone. India and Macromalayana are inhabited 
by the same, together 6 species. Sumatra and Borneo have the same number, except a Taraka which did not 
pass over to Borneo. Java possesses yet five species, Lombok only three. 
In Celebes there are seven species, among them an autochthonous one; New Guinea has three, all of 
which are endemic. The Aru and Key Islands are not inhabited by the Castaliinae, whereas four species have 
reached the Philippines. Formosa is only inhabited by one species. Interesting is the aboundance of Celebes 
in contrast with Java; this fact is naturally explained by the circumstance that the oriental caleta has immi¬ 
grated there from India by way of the large Sunda Islands — which species, as far as is known, reaches neither 
the Malayan Peninsula nor any of the three large Sunda Islands, besides Celebes possesses endemic forms in 
C. ilissus and fasciatus. 
The single species are distributed as follows: 
