MAMMALIA. 37 



chiefly on the offal of the turtles, slaughtered for the daily food of the inhabitants. 

 If the settlement were destroyed, I feel no doubt that this latter variety would be 

 compelled to migrate from the coast. Did it originally descend from the summit? 

 and, in the case just supposed, would it retreat there? and, if so, would its black 

 colour return ? It must, however, be observed, that the two localities are sepa- 

 rated from each other by a space, some miles in width, of bare lava and ashes. 

 Does the summit of Ascension, an island so immensely remote from any conti- 

 nent, and the summit itself surrounded by a broad fringe of desert volcanic soib 

 possess a small quadruped, peculiar to itself ? Or, more probably, has this new 

 species been brought, by some ship, from some unknown quarter of the world ? 

 Or, I am again tempted to ask, as I did in the case of the Galapagos rat, has the 

 common English species been changed, by its new habitation, into a strongly 

 marked variety?" — D. 



Mr. Darwin seems to have foreseen the difficult problem which these two rats 

 have furnished, and although I have spent much time in studying the Muridse, I 

 must confess I have been exceedingly puzzled by the animals in question. It 

 appears as if the brown, and black rats, (M. clecumanus, and M. Rattus,) and 

 likewise the common mouse, (M. Musculus,) * all of which follow man in his 

 peregrinations, and which, to a certain degree, are dependent upon man, and may 

 therefore be termed semi-domestic animals ; like really domestic animals, are sub- 

 ject to a greater degree of variation than those species which hold themselves 

 aloof from him. 



Upon the whole then I have determined to describe the two Ascension 

 Island specimens as one species, and as varieties of the Mus Rattus, but with 

 a mark of doubt, since I do not possess sufficient materials for a rigorous ex- 

 amination, having, in fact, but one skin of each variety, and neither skull nor 

 skeleton. I have also applied the name of insularis, to designate this variety 

 or species, whichever it may be, for, supposing it be not a distinct species, it is 

 so marked a variety, that a name for it is desirable. 



* The great Bandicoot rat of India, {Mus gigantens, of Hardwicke,) ought, perhaps, to be added to the 

 species above enumerated ; and I strongly suspect several catalogued species will prove but varieties of this 

 animal. 



