e compare it with some other land masses that may be termed 'islands.' Geology teaches us in fact that

slands, which are more remote in their position from continents and concerning which it seems clear that they have originated de novo by the action of submarine volcanos or of the growth of coral, combined with subsidence, following elevation, or from several of the causes combined. In any case the islands which are termed oceanic islands have never formed part of a continent. They are not relics of previously existing continents. It become

g's Hypogaeon havaicum), A. foetida, A. caliginosa, A. nordenski?ldi, A. limicola, A. rosea, and finally the well-known Pontoscolex corethrurus. Of these species there is only one which is even possibly a form limited to the Sandwich Islands, and that is Pheretima perkinsi, a species which I myself at first described as a new form, but which was afterwards regarded

soma, D. helophila, D. perionychopsis among the Megascolecidae, besides Phreodrilus campbellianus, Pelodrilus tuberculatus, P. aucklandicus and the Lumbricid Helodrilus constrictus. There were also four species of purely aquatic Oligochaeta which we shall leave aside from the present enumeration, as their range in space is a matter requiring a different explanation from that of the terrestrial forms. Here we have a series of worms,

elsewhere, but the majority of them do not consist of widely ranging peregrine forms. It appears therefore most probable that these islands are not oceanic islands but a portion of the former existing northern portion of the