1004 VARIATION 
perceptible to the carefully-trained eye! It isa trite remark that ina 
flock of sheep the ordinary man sees nothing to distinguish one animal 
from another, while the shepherd knows each unfailingly, and those 
who look after birds kept in captivity are soon able to do the same 
in regard to their charges, though both shepherd and bird-keeper 
would often find it impossible to point out wherein the difference 
lay. Yet because the difference cannot be expressed in words, its 
existence is not to be denied, and indeed for all practical purposes 
we may assume its existence except in rare cases. ‘Thus, believing 
Variability to be general, the question naturally arises as to its 
limits, if it has any. Some there are who would boldly declare it 
to be in one sense boundless, and others would define its limits as 
geographical. Much is to be said for this last point of view, which | 
was that taken by some of the earliest investigators of Variation ; but 
then it must be admitted that those who adopt it have a very summary 
way of treating the subject, though it is eminently practical and 
perhaps at present indispensable. 
When a definite structure or coloration of any form is observed 
to be associated with a definite area, and to cease on that area being 
overpassed, the systematist will generally say at once that we have 
two distinct species of the form; but, on the other hand, he is often 
puzzled how to regard a form that ranges over an area, mostly 
large, at one end of which it exhibits certain well-marked characters 
which gradually vanish as the centre is approached, and are as 
gradually replaced by different but equally well-marked characters, 
until at the other end it generally, though not always, assumes a 
wholly distinct appearance, the intervening space being thus occupied 
by individuals more or less intermediate between the extremities of 
the series. The reflective naturalist will perceive the probability of 
both these categories being reducible to the same principle—only 
in the latter case the Variation is continuous and in the former 
discontinuous; but it has taken ornithologists a long while to 
recognize this probability. 
Conditions such as these are furnished in cases far too numerous 
to name,” and their proper recognition and full appreciation (if that 
specifically characteristic, so that it was utterly impossible to declare the limits 
of individual modification, while the Variability did not wholly depend on age 
or sex (Phil. Trans. 1869, pp. 830, 831). Subsequent examination of a still 
larger series of specimens confirmed the former statement (op. cit. vol. 168, p. 
451). Yet all this amount of Variation was exhibited by the individuals of a 
single species, confined to a small island, and apparently all living at about the 
same period and under the same conditions. 
1 The more minutely a specimen is described the less chance there is 0 
another specimen being found to agree with it. Hence the value of a proper 
diagnosis, in the old sense of the word, compared with a description—a fact 
which some modern ornithologists are apt to overlook. 
* They are mostly found among the Oscines, but possibly because that group 
