TRANSACTIONS OF SECTION D. 789 



remain in the centre of distribution -we find tlie more decidedly distinct species at 

 tlie extremes of the range, caused not by interbreeding, but by differentiation. 

 To illustrate this by the group of the blue titmouse. AVe find in Central Russia, 

 in the centre of distribution of the family, the most generalised form. Partis pleskii, 

 partaking of the characters of the various species east, west, and south. In the 

 north-east and north it becomes differentiated as P. cyanevs ; to the south-west 

 and west into P. Cfcruleiis and its various sub-species, while a branch extending 

 due east has assumed the form of P. Jlnvipectus, bearing traces of affinity to its 

 Keiglibour P. cyaneus in the north, which seems evidently to have been derived 

 from it. 



But the scope of field observation does not cease with geographical distribu- 

 tion and modification of form. The closet systematist is very apt to overlook or 

 to take no count of habits, voice, modification, and other features of life which 

 have an important bearing on the modification of species. To take one instance, the 

 short-toed lark {Calandrella hrachydacfyla) is spread over the countries border- 

 ing on the Mediterranean ; but, along with it, in Andalusia alone is found another 

 species, C. bcvtica, of a rather darker colour, and with the secondaries generally 

 somewhat shorter. "Without further knowledge than that obtained from a com- 

 parison of skins, it might be put down as an accidental variety. But the field 

 naturalist soon recognises it as a most distinct species. It has a different voice, a 

 differently shaped nest ; and, while the common species breeds in the plains, this 

 one always resorts to the hills. The Spanish shepherds on the spot recognise their 

 distinctness, and have a name for each species. Take, again, the eastern form of 

 the common song-thrush. The bird of North China, Turdiis auritus, closely 

 resembles our familiar species, but is slightly larger, and there is a minute differ- 

 ence in the wing formula. But the field naturalist has ascertained that it lays 

 eggs like those of the missel-thrush, and it is the only species closely allied to our 

 bird which does not lay eggs of a blue ground colour. The hedge accentor of 

 Japan (Accentor rubtdus) is distinguished from our most familiar friend. Accentor 

 modularis, by delicate differences of hue. But, though in gait and manner it 

 closely resembles it, I was surprised to find the Japanese bird strikingly distinct in 

 habits and lile, being found only in forest and brushwood several thousand feet 

 above the sea. I met with it first at Chiusenze — 6,000 feet — before the snow had 

 left the ground, and in summer it goes higher still, but never descends to the 

 cultivated laud. If both species are derived, as seems probable, from Accentor 

 immaculatus of the Himalayas, then the contrast in habits is easily explained. The 

 lofty mountain ranges of Japan have enabled th.e settlers there to retain their 

 original habits, i'or which our humbler elevations have afforded no scope. 



On the solution of the problem of the migration of birds, the most remarkable 

 of all the phenomena of animal life, much less aid has been contributed by the 

 observations of field naturalists than might reasonably have been expected. The 

 facts of migration have, of course, been recognised from the earliest times, and 

 have afforded a theme for Hebrew and Greek poets 3,000 years ago. Theories 

 which would explain it are rife enough, but it is only of late years that any 

 systematic eff'ort has been made to classify and summarise the thousands of data 

 and notes which are needed in order to draw any satisfactory conclusion. The 

 observable facts may be classified as to their bearing on the Whither, When, and 

 How of migration, and after this we may possibly arrive at a true answer to the 

 "Why ? Observation has sufficiently answered the first question, Whither ? 



There are scarcely any feathered denizens of earth or sea to the summer and 

 winter ranges of which we cannot now point. Of almost all the birds of the holo- 

 arctic fauna, we have ascertained the breeding-places and the winter resorts. Now 

 that the knot and the sanderling have been successfully pursued even to Grinnell 

 Land, there remains but the curlew sandpiper (Tringa S2tbarqt(atci), of all the 

 known European birds, whose breeding ground is a virgin soil, to be trodden, let 

 us hope, in a successful exploration by Nanaen, on one side or other of the North 

 Pole. EquaUy clearly ascertained are the winter quarters of all the migrants. 

 The most casual observer cannot fail to notice in any part of Africa, north or 

 eouth, west coast or interior, the myriads of familiar species which winter there. 



