136 DR. J. STEPHENSON ON THE MORPHOLOGY, CLASSIFICATION, 



tolerably calm, such a raft might be carried along by the current, 

 aided by the wind acting on the trees, till after a passage of several 

 weeks it might arrive safely on the shores of some land hundreds 

 of miles away from its starting-point." 



Overseas colonization is a very remote chance, it is true, in any 

 given length of time; but, says Matthew, if we nudtiply the 

 almost infinitely small chance that such colonization takes place 

 in any given length of time, such as a year, by the almost infinite 

 duration of geological periods, we obtain a finite and quite 

 probable chance. For example, the time during which natural 

 rafts have been observed covers about three centuries, while the 

 duration of Casnozoic time is estimated as three million years ; 

 if we allow that ten cases of natural rafts have been recorded 

 during these three centuries (the wording of the extract from 

 Wallace given above would, however, seem to indicate that this is 

 an understatement), a thousand may have actually occurred in 

 this time, and hence thirty million in the whole Camozoic (this is 

 a miscalculation — it should be ten million). He then makes 

 certain assumptions regarding the occurrence of living mammals 

 on such rafts — as to the chances of there being a couple, or a 

 gravid female, and as to the dangers of landing ; and his con- 

 clusion is that the number of cases during the Csenozoic in which 

 mammals will have established themselves on the larger oceanic 

 islands is of the order of 300 — quite enough at any rate to cover 

 the dozen or two known cases. With invertebrates the chances 

 would be much greater. 



And certainly, whatever the possibilities of the transfer of 

 mammals by rafts, the transfer of earthworms must be far more 

 probable. Such rafts as have been described above may or may 

 not bear mammals — Matthew's calculations are based on the 

 supposition that they do so only once in a hundred times ; but 

 every one will probably contain earthworms, in the soil, under 

 the bark of living trees, in the axils of their leaves, or in rotting 

 wood. Nor are worms restricted to the larger rafts ; the smaller 

 worms of euryhaline groups (those that can withstand saltwater) 

 and especially their cocoons, may probably be transported for 

 long distances in masses of tangled seaweed ; Michaelsen, himself 

 a bridge-builder, presses this point against Benham in explaining 

 the distribution of Microscolex in the Subantarctic regions (13). 



It may be asked, too, whether earthworms are in general so 

 readily killed by salt water as is assumed. It is well known 

 that many Enchytrseids and Tubificids are regularly found on the 

 shore ; and among the higher groups the genera Pontodrilus, 

 Pontoscolecc, and Microscolexha,ve the same habitat often, though not 

 always ; I have received HoplocJuntella from the shore of western 

 India, though the genus Avas not previously known from such 

 localities. It is at least possible that many worms are capable 

 of speedy acclimatization to salt, just as a fresh-water Amoeba 

 can be acclimatized by the gradual addition of salt to its water. 

 And it is remarkable how difficult it is to come at any definite 



