Vil THE ANTARCTIC AND CIRCUMTROPICAL ZONES 201 
mysis scyphops, which occurs at both poles, but is not known from 
the tropics. This is a pelagic species, and we know that the 
Mysidae often descend to considerable depths. We also know 
that the Mysidae are dependent on cold water, only occurring in 
boreal or temperate waters. We may safely suppose, therefore, 
that the migration of this species has taken place by their for- 
saking the surface-waters as the tropics were approached, and 
passing down into the depths where the temperature is constantly 
low even in the tropics. 
The dependence of Crustacea upon the temperature of the 
water is also illustrated by the distribution of the Lithodinae. 
The headquarters of this family are in the boreal Pacific, with a 
few scattered representatives in the boreal Atlantic. The cool 
currents on the western coasts of America, however, have per- 
mitted certain forms to migrate as far south as Patagonia, where 
they still have a littoral habit. In the tropical Indo-Pacific, 
where a few species occur, they are only found in deep waters. 
Thus at these various latitudes, by following cool currents or 
migrating into deep water, they are always subjected to similar 
conditions of temperature. The same kind of thing is observed 
in Arctic seas, where deep-sea forms are apt to take on secondarily 
a littoral habit owing to the temperature of the depths and of 
the shore being the same. 
Despite the impassable barriers of land which now sever the 
tropical oceans, we can yet speak of a circumtropical zone 
possessing many species Common to its most widely separated 
parts. Such circumtropical species, occurring on both the Atlantic 
and Pacific coasts of tropical America, on the West African coast, 
and in the Indo-Pacific, are various Grapsidae, Calappa granulata 
and its alles, and certain Albunea. The most striking instance 
of all is that of the Land-crabs. Of Ocypoda, the greater number 
of species occur in the Indo-Pacific, but representatives are also 
found on the tropical Eastern and Western American coasts and 
on the West African coast, and the same is true of Gelasimus. 
The genus Cardisoma, belonging to a different group of Land- 
crabs, is also typically circumtropical. 
For this community of the circumtropical species we may 
certainly advance in explanation the comparatively recent forma- 
tion of the Isthmus of Panama. Besides the resemblance of the 
Crustacea on the east and west coasts of the isthmus, we have an 
