10 VARIATION, DISTRIBUTION, AND EVOLUTION OF THE GENUS PARTULA. 
exposed reefs and the volcanic materials may well cover a far more ancient massif 
of basic nature, whose subsidence so altered the isostatic conditions as to bring 
about the partial compensation of volcanic outflow. 
Turning to the biological considerations, it must be emphasized that the very 
fact of the present isolation of the Mariana Islands, especially on the southward 
side, strongly supports the subsidence theory, which at first sight it would appear to 
confute. If the islands in question had arisen de novo, then the Partule now in- 
habiting them could not be considered as indigenous, in the sense that the species 
of the Society Islands are held to be original. It would be necessary to postulate 
an immigration into such areas from other islands bearing Partule at the time. 
We know from the situation in the Society Islands that the snails of one island are 
not transported to a nearby neighbor in the same group, either by floating rafts or 
by human agencies. Hence, a fortiort, we would not expect oceanic drift to bear 
new settlers to the Mariana Islands from other groups, such as the Caroline Islands, 
when the degree of isolation is as great as it proves to be. Despite the geological 
difficulties, the biological findings strongly support the view that the dominant 
process in this part of Oceania has been one of subsidence and of insular dissection. 
It may be that further study of this problem will alter the author’s present 
conclusion; however, the original evolutionary history of the Mariana species of 
Partula is not the subject of discussion at this time, for we are here concerned with 
the detailed analysis ot the Partula inhabitants within the confines of the group as 
it now exists. The problem of external relations may well await its due considera- 
tion when additional studies of specific island associations of this genus shall have 
been completed. 
CURRENTS, WINDS, AND STORMS. 
As in the volume on the Tahitian species of Partula, the position is taken that 
the winds and currents have not been active in any way or degree as factors of the 
dispersal and distribution of the forms now under discussion. It is true that one of 
the Guam species, P. gibba, occurs also in Tinian and Saipan, and possibly in Rota, 
and that these islands lie in a direction which is to the leeward of Guam during 
certain months of the year. It is therefore conceivable that masses of vegetation 
could be carried from Guam to the neighboring islands by winds and currents, there 
to be stranded upon the shores; in such cases, were snails adhering to the leaves and 
branches of the floating rafts, they might become the first settlers in a previously 
uninhabited area. But it is not believed that the conditions for such an episode 
are fulfilled; the winds are only indirect influences, not on the dispersal and inter- 
island transport of the snails, but upon the amount of rainfall and the degree of 
humidity, which factors directly condition the mobility and feeding habits of the 
animals in a given suitable region of vegetation on an inhabited island. 
The winds vary in their directions throughout the year, in ways that are 
precisely indicated by the figures for an illustrative period, 1902, as given in Safford’s 
invaluable compendium (table 1). In general, it appears that a trade-wind of 
considerable power and regularity blows from the northeast and east during the 
greater part of the year. This wind becomes less steady with the advent of June; 
