56 VARIATION, DISTRIBUTION, AND EVOLUTION OF THE GENUS PARTULA. 
island inhabited by the species, some of the valley colonies are so low in vitality as 
to be in imminent danger of extinction. It is indicated as probable, therefore, that 
the former range of hyalina extended over the whole Society Group, and even so 
far as to include the high islands of the Austral and Cook Groups. ‘The facts in 
the case of Partula attenuata are extremely significant in this connection, for that 
species lives in Raiatea as well as in Tahiti, but not in Huahine, Tahaa, or Moorea. 
It occurs only in the remotest and highest portions of the valleys, on the upper 
portions of trees quite different from the plantain and other nurse-plants. It can 
not be argued that a form with such habits could be introduced by human agency 
into Raiatea from Tahiti or vice versa; it must be a species which had a far wider 
distribution at a remote period, and which has disappeared everywhere save in the 
two islands specified. 
The greater obstacle to the acceptance of the general interpretation with which 
we are dealing consists in the limestone character of Mangaia and Moki. These 
are uplifted atolls devoid of volcanic centers. According to the doctrine of sub- 
sidence, they are formed out of coral constructed about and above a volcanic cone, 
which has entirely withdrawn or has weathered away to a level below that of the sea. 
A subsequent uplift of the floor of the ocean has brought the limestone mass to its 
present position. If hyalina had existed upon the original mountainous island, the 
reduction of the land surface to sea-level would presumably have destroyed the 
environmental conditions under which the animals could live; if it could not have 
withstood the relatively dry atmosphere of a low atoll, it must have become extinct 
before the reverse geological process of upheaval had reestablished a higher region 
with dense vegetation and shade like that now occupied by the animals in Moki 
and Mangaia. Under such circumstances the present occurrence of hyalina in these 
places would therefore be due to human introduction. 
In my opinion, however, it is not necessary to abandon the consistent view that 
the present colonies of hyalina, existing on widely separated islands of diverse 
geological nature, are the remnants of an ancient species with a former wide range 
in Polynesia. In discussing the facts in the case of Tahitian colonies, it was shown 
that hyalina thrives in the drier quadrant of the major element, in the wider and 
more open valleys of all quadrants, and in the less moist situations within the valley 
areas of its occurrence; as a species, it shows a marked tolerance of heat and drought. 
Hence it does not seem unreasonable to suppose that highly resistant individuals 
might survive the vicissitudes of existence during the geological transformations 
involved in the production of an uplifted limestone island like Mangaia and Moki. 
In conclusion, then, Partula hyalina seems to be an old species specialized to 
such an extent that pigmentation has been completely lost. Subsequent alterations 
of its originally continuous territory, through which island groups and islands were 
isolated, led to its dispersal in separate island colonies on volcanic peaks. From 
nearly all of these it has disappeared, owing to an inherent lack of vigor in certain of 
its associations, while even in Tahiti, where it still exists, it has disappeared from some 
of the valleys for the same reason. The shells vary colonially in dimensions, in 
relative measures, and in their combinations of absolute and relative characters. 
