


1891.] Geography and Travel. 903 
Barrington (eight specimens), South Albemarle (one specimen), Dun- 
can (one specimen), East Albemarle (one specimen) ; on Chatham one 
was observed. There cannot be any doubt that this mammal is an 
original inhabitant of the group. 5 
The birds are exceedingly interesting, and I hope to be able to give 
a satisfactory solution of the Geospiza question. As I have shown in 
a paper published in the Biologische Centralblatt, the Iguanoid ` 
Tropidurus is represented by a single species on each island, and 
nearly every island contains a different species or race of Tropidurus. 
This has been absolutely sustained. Now in the plastic genera of birds 
we find exactly the same. Let us first consider the genera Nesomimus 
and Certhidea, of which only a single species is found on each island. 
The genus Nesomimus is represented by a different species or race on 
every island, and there is never more than one species or race found 
on one island. The same is true of Certhidea, but this genus is not 
quite so plastic. On Hood, as it is known, Nesomimus is much differ- 
ent from the other species ; but so are Certhidea and Tropidurus. On 
the central islands—Chatham, Indefatigable, Jervis, James, Albe- 
marle—Nesomimus shows only slight differences; but so do Certhidea 
and Tropidurus. What I want to state is the absolute harmony in the 
distribution and the grade of difference in these forms. This is also 
true of the flora, as far as I could make out. 
It is now necessary to examine such genera as Geospiza, Camarhyn- 
chus, and Cactornis, which are represented by a greater number of | 
species on one island. As it is well known, there has been a great 
uncertainty as to the number of species found one ach island ; on some 
islands not less than eight species have been recorded. 
So far asmy present investigation reaches, probably on none of the 
islands visited is the number of species greater than three; but these 
three species vary nearly on each island, each separately, as if they 
would represent three different genera. The same view I have for 
Camarhynchus and Cactornis, and all such genera which contain more 
than one species of a genus on a single island,—like Bulimus, for 
instance. But often we find that the highest number of species is not 
reached by every island, but that the number of species is reduced. 
This can be explained by the extinction of one of the species. In this 
respect I have first to make some remarks about Nesomimus. As is 
known, Nesomimus existed on Charles Island in 1835 when Darwin 
visited the island ; it still was found there in 1868 during Dr. Habel’s 
visit. We did not see a single specimen on Charles Island, notwith- 
HRA the whole island was crossed and three different ports were 
at.—October.—4. 

