Birds of Celebes: Psittacidae. 
167 
According to the above methods of reasoning, therefore, the genus Loriculus 
consists of two branches and one aberrant species of uncertain position, L.gal- 
giilus. The two branches by their young throw back to a form resembling 
^L. vernalis of the Indian countries, and the more highly-differentiated species of 
the southern branch by their females and immature males throw back to the 
simpler species of their branch, which seem to stand nearest to L. vernalis. 
L. vernalis should, therefore, stand nearest to the ancestral type. From the 
postulate, that colonists get most changed, the Indian countries should, therefore, 
be the original habitat of the ancestral type, which became more and more 
changed as it proceeded to the Philippines on the one hand, and to Java, 
Celebes, Halmahera, New Guinea, on the other. 
The alternative explanations should now be examined. It may be argued 
in the first place that the reverse of all the above process has taken place, — 
that the complex species of New Guinea, Halmahera, and Celebes were the 
original forms and extended their range westward, gradually losing their speci- 
alizations of plumage until they became simplified into the modest-looking 
L. vernalis of India. There is the more reason for advancing this argument, as 
a retrograde development, or at least a change from more complex differentia- 
tions of structure and colour to simpler ones, may certainly take place. Species 
may have a culminating point set to their evolution beyond which they cannot 
develop, and having attained this they turn again into less specialized forms 
till they die out, — not from external causes, but because their vitality is ex- 
hausted and their career is over. As it is with the individual, so it may be 
with the species, the genus, the family, and so on. The very centre of the 
parrots in the Old World is at the present time (and may always have been) 
Australia and the eastern part of the East Indian Archipelago ; Wallace looks 
upon the Oriental region as the land of origin of the Psittacidae., but this is 
far from being proved in any way, and Reichenow, for instance, offers reasons 
for making the hypothesis of the eastern origin just as plausible. At all events 
one has in either case to deal with mere probabilities, as proofs will be want- 
ing till palaeontological facts allow men to abstain from speculations. (Compare 
Furbringer’s remarks in his fundamental “Untersuchungen” pp. 1116, 1287 and 
1293). If the “Stammform” of Loriculus originated in New Guinea or the 
neighbourhood and spread from there westward, the ontogenetic characters, 
which are treated above as remnants of the phylogenetic development of the 
genus, must have another meaning, which we cannot explain. The argument 
that Loriculus spread its range from Papuasia to India, though it has this in 
its favour that the genus issued from the countries which are the richer in 
Parrots, fails, however, to account for the existence of the branch found in 
the Philippines and Ceylon, in which the ontogenetic development differs from 
that of the southern branch to which the Papuan species belong. 
The third alternative explanation is that Loriculus ranged from India to 
