189 
Mr. H. Seebohm on the Genus Scolopax. 
each species the range of variation is so large that the con¬ 
struction of hard-and-fast diagnoses is very difficult. The 
Semi-Woodcocks naturally divide themselves into two groups: 
half the species having the inner and outer webs of the first 
primary uniform in colour, whilst in the other half the outer 
web is white or barred with white. The same character 
appears to be also an important one in the typical Snipes; 
and, so far as those which inhabit Europe, Asia, Africa, and 
North America are concerned, it appears to be a fairly con¬ 
stant one; but in South America it breaks down altogether. 
In the four subspecific forms or local races of Scolopax 
frenata the outer web is sometimes white, sometimes brown, 
occasionally white on the basal half and brown on the ter¬ 
minal half’, and examples occur in which it is white on one 
wing and brown on the other. The number of tail-feathers 
is also so variable in many, perhaps in most, of the species 
that its practical value as a diagnostic character is very small 
indeed. The extent to which the outer tail-feathers are 
barred does not vary quite so much as the colour of the 
axillaries in some species; but the shape of the outer tail- 
feathers appears to be slightly more reliable. There can be 
little doubt that the diagnosis of one or more tail-feathers on 
each side stiffened and attenuated , though somewhat vague, 
would exclude the five Woodcocks and the four typical 
Snipes yet to be considered, whilst it would include the seven 
Semi-Woodcocks and the twelve typical Snipes already 
considered. The importance of the outer tail-feathers as a 
specific character in this genus becomes obvious when the 
somewhat startling fact is realized that, if the outer four or 
five feathers on each side of the tail were cut away in ex¬ 
amples of S. gallinago, S. wilsoni, and S. paraguayce , it would 
be impossible to guess to which of the three species any one 
of them belonged. 
It may appear an arbitrary and clumsy arrangement; but 
it is necessary to draw some definite line which will diagnose 
the four typical Snipes yet to be considered from the twelve 
preceding species. The two monster Snipes of South Ame¬ 
rica may be dismissed on account of their barred primaries, 
l 2 
