1897.] Birds of the Gulapagos Archipelago. 779 
I shall show now that the Cactornis propinqua Ridgway from 
Tower Island in the north and the Geospiza conirostris Ridgway 
from Hood Island in the south of the Archipelago have no re- 
lationship whatever. The distance between Hood and Tower 
is one hundred nautical miles. Mr. Ridgway arranges the 
different species of Geospiza (including Cactornis) in a single 
line, to show the gradual connection between the different 
forms. I can not agree with this. 
My opinion is the following: All the plastic genera, which 
are represented only by a single species on each island, as 
Nesomimus, Certhidia, Pyrocephalus and Cactornis, show pecu- 
liar species on nearly every island. The same is true for the 
iguanoid lizard Tropidurus, the land tortoises, for Phyllodacty- 
lus of the Geckonidz, of the genus Schistocerca of the Orthop- 
tera,” and one of the most striking examples is offered by 
Euphorbia viminea Hook. fil. of the Euphorbiacee.* 
But there ‘are genera, like Geospiza and Camarhynchus, 
which have more than one species on one island—two or 
three, perhaps four.t How can we explain this? I think it is 
not difficult to answer this question. We simply have to im- 
agine that already, before the splitting up of the Galapagos 
landarea into distinct islands, there existed at least 3 species 
of Geospiza and Camarhynchus, each of which became differ- 
entiated on the different islands. This shows at once that we 
can not arrange these species in one series, but in 3 parallel 
series. In some islands, as will be seen from the table, one or 
the other form of the 3 series may be missing. This conclu- 
*Baur,G. The differentiation of species on the Galápagos Islands and the 
origin of the group. Biological lectures del. at the Marine Biol. Laborat. of 
Wood’s Holl, summer session of 1894. Boston, 1895, p. 67-78. 
* Robinson, B. L. and J. M. Greenman. On the Flora of the Galápagos Islands, 
as shown by the collections of Dr. G. Baur. Amer. Journ. Science, Vol. L, Aug- 
ust 1, 1895, p. 135-149. 
tI consider Geospiza magnirostris Gould and G. strenua Gould as unseparable 
Species which are based on a single specimen like Geospiza dentirostris Gould, and 
the locality of which is unknown, are of no use. Geospiza difficilis Sharpe must 
be restricted to Abingdon. This leaves 3 species only for Charles. If there are 
more than 3 or 4 species of Geospiza said to occur on one island, I have no doubt 
that there is some mistake. 
