882 
CASTALIUS. By H. Fruhstorfer. 
phamus. from South Celebes. Unknown to me in nature. — pharnus Fldr. ( = epius Holl *). Described from Amboina 
by Felder, found in Burn by Doherty, and as a novelty for the Key Islands from Key-Tual in my collection. 
Semper already in 1889 stated Buru as the habitat, Smith in 1894 Gani in Halmaheira. Specimens considerably 
strigotus. smaller than those of titius, under surface lighter greyish-white. — strigatus Semp. flies in October. Described 
from Cebu by Semper, in a very closely allied form as a novelty for the island from Palawan in my collection. 
georgi. strigatus will of course be yet discovered on all the southern Philippines, particularly in Mindanao. — georgi 
Fruhst. Bohol. A very different race owing to the uncommonly light colouring and the blurred longitudinal 
lines beneath, perhaps a dry period form of the preceding, although the flying-time being stated as November 
sett)peri, scarcely differs from that of the preceding race. — semperi Fruhst. This insular form mixed up by Semper 
with that from Celebes is already separated from Snellen’s form by Semper’s strigatus. But if the Central 
Philippines such as Cebu and Bohol already produce a race differing from the Celebic race, the far more nor¬ 
thern Luzon can impossibly be inhabited by a form more closely allied to substrigata, or even by substrigata 
dilama, itself. For the Luzon-race therefore the name semperi is reserved. — dilama Mr. (154 f as didama) is described 
by Moore to be beneath darker brownish-grey and with fainter zigzag-lines than epius from India. The white 
transcellular spot of the forewing is larger. The latter is also the case in the Formosan and as the Formosan 
<§E also correspond with Moore’s statements, I let the name dilama temporarily be current for the Formosan 
form. Hainan, Formosa. According to a statement of de Niceville **), the pupa is only suspended on a 
cremaster-thread, without a median band as in the other Lycaeninae which also repeats itself in Aphnaeus, 
Tajuria, Cheritra, Horaga and Loxura. Doherty ***) writes that the eggs of Spalgis are flattened off above 
and finely covered with irregular hexagons. The position of Spalgis can only be understood when the lepi- 
doptera of tropical Africa, where such great numbers of low Lycaenid forms are stored will once be better 
known. 
13. Genus: Castalius Hbn. 
This small bi-continental group of species, scarcely comprising 10 certain species in Indo-Australia, 
has a very short history. Its species were mostly discovered in the middle of the last century and published 
under the collective name of ,,Lycaena‘\ being customary at that time. But already in 1869 Butler ascer¬ 
tained that Hubner had introduced a special name, Castalius, for a species having been already described 
in the eighteenth century. Butler accepted Hubner’ s transcription and logically assigned the name Castalius 
in the Cat. Fabrician Diurn. Lepidopt., p. 169, to the only species Castalius rosimon known to the ancient authors 
(Fabricius, Cramer, Hubner). This name of Hubner-Butler was then no more altered for half a century 
- which case is perhaps quite unique among the names of the Lycaenid genera having so often been dug out 
and in. Distant in 1884 even established a subfamily ,,Castalaria“ ; a systematic monstrum, it is true, since 
it contains heterotypical genera such as Nacaduba and Everes, but does not comprise entirely genuine ,,Castalaria“ 
such as Taralca. — Of the eminent modern authors only Aurivillius suppressed the name Castalius and 
distributed the Ethiopian species on three groups of his collective genus ,,Cupido“; which proceeding is scienti¬ 
fically scarcely contestable, but it forces us to keep in memory transcriptions such as ,,seventh, ninth and 
tenth group”, whereby a much less clear mnemotechnical constellation is created than if we adhere to the nice 
name ,,Castalius”. And although the ,,genus Castalius “, in spite of all the modern attempts to lay hold of it 
in a morphological way (androconia, anatomy), has not yet offered us any prehensible mark at all, it still reckons > 
among those Lycaeninae-groups that are at once recognizable without any structural characteristic mark, 
only b\ their facies. The chief character of all the Castaliids is a larger or smaller accumulation of black spots 
on both wings beneath occurring only with them. Most of the species exhibit besides two eye-spots surrounded 
by a metallic lustrous crescent (alliance with the Tarucinae). Some species do not. possess any ocelli (alliance 
with Lycaenopsis). All the Castaliinae , however, whether with or without ocelli, are to be separated from all 
the other Lycaeninae by the invariably purely white base of the costal region on both wings, which is besides 
separated by an almost invariably straight subbasal stripe or spot from the median area being more or less 
speckled black. In some insignificant Thysonotis- species there is also a white base of the wings, but the sub¬ 
sequent black streak is expanded like a band. By reason of the white basal region of the subsequent stripe we 
are able to ascertain most accurately the homogeneousness not only of all the Indian but also of all the African 
Castaliinae. But if we take in view the structure of the veins of the wings, we do not find a single tenable 
*) Nov. Zool. 1900, p. 67. Buru. 
**) Journ. As. Soc. Beng. 1900, 247 
***) Ibid. 1889. 
