OF THE GALAPAGOS ARCHIPELAGO. 469 
Jervis Island should also be visited, though they are little else but outlying rocks of 
the larger islands, James and Indefatigable. 
Lastly, of unvisited islands there remain Wenman and Culpepper. The chance of 
finding any land-birds on either of these rocks is exceedingly small; but the fact of 
their non-existence should be ascertained. 
All the islands that have already been investigated are worthy of further exploration. 
The species of Mimus require more working out. Several of the so-called species of 
Geospiza are but little known, such as G. magnirostris, G. nebulosa, and G. dubia; 
also more specimens of Cactornis assimilis, C. abingdoni, and C. pallida are required to 
complete our knowledge of these birds. 
Camarhynchus psittaculus and C. crassirostris are but little known, and still less 
the bird described by Lafresnaye as C. cinereus. ‘The two Herons (Butorides plumbea 
and Nycticorax pauper) are as yet only known from immature birds; more specimens 
of the Flamingo would be useful, and a good series of the Petrels. Lastly, and 
most important of all, a series of observations are required upon the attitude the 
different species of Geospiza maintain towards one another, tending to show how far 
the differences observable, or thought to be observable, in dried specimens indicate 
the actual grouping in species of living individuals. 
From these remarks it will be seen that much useful work remains to be done in 
these islands in the birds alone. In other branches of biological science even more 
investigation is required. 
VY. ON THE VARIATION OF THE SPECIES IN CERTAIN GENERA, AND THE CONSEQUENT 
DIFFICULTY IN DEFINING SpeEctFic Limits. 
The acquisition of a large series of specimens of the different so-called species of 
Geospiza renders the question as to the definiteness of the points of specific distinction 
between them more difficult of solution than ever. Distinctions are plainly enough to 
be seen between such birds as G. magnirostris and G. parvula, where great disparity 
in size is obvious. But these differences are gradually erased by almost insensible 
steps by the interposition of G. strenua, G. fortis, and G. fuliginosa. The series 
before me includes specimens that can almost as well be referred to either of two 
contiguous species, so that their position can only be determined by assigning to each 
species what must be called arbitrary standards of measurements of the bill alone. The 
question follows, Do these birds, in their natural relations to one another, keep them- 
selves to groups of individuals dependent upon the size of their bills? The answer is 
to be found rather by field-observation than by measuring skins. Were different 
species, though closely allied, found to inhabit different islands, the case would be 
much simplified. But what do we find? Charles Island possesses three out of eight 
so-called species, Chatham Island has no less than seven, James Island four, Indefati- 
