( 74) 



Not only the birds of onr northern latitmles are mijiTauts, but also manj- of 

 the insect-eating birds of the Atlas Mountains and North Algeria leave their homes 

 dnring the winter, sncli as the Nightingales, Redstarts, Warblers, Swallows and 

 others. None of these are pnshcd farther south by the inrush of nortiiern 

 migrants ; neither Sylvin deserticola, Diplootociix moussieri nor Sj/lciiy cantHlauR 

 inornata and others have ever been found in Tropical Africa: they leave the 

 mountains in the winter and pass the cold season in the Northern Sahara, while 

 the arctic birds pass throngh their hanuta ami proceed far(her southwards. In 

 the spring we see the northern birds pass through, when the Algerian ones are 

 already settled and begin nesting I 



AVe should be able to prove more snch facts if we could better recognise the 

 local races, and I do not doubt that the enormous strides we are making in 

 distinguishing the various subspecies will bring more such facts to light ; in some 

 cases, where we are unable to separate the birds from widely different countries, 

 only " bird-marking " may solve such (jnestions. 



We have hitlierto been accustomed to look upon, as breeding in N.W. Africa, 

 all birds regularly found there dnring the second half of Jlay and in June. Our 

 observations of 1912 — partly confirmed by Mr. Spatz — show that such conclusions 

 mnst be drawn very slowly and carefully. A jierusal of the foregoing pages will 

 show that wc have met numerous undoubted migrants in the Sahara after the 

 middle of May and even in the month of Juno I There is no reason whatever 

 to doubt tliat this late migration is a regular phenomenon. As some of these 

 late migrants were observed while their brethren in the north began already to 

 nest, there is reason to suppose that some of them were not breeding in that year, 

 some apparently because t hey are too young or too old, or — in some cases certainly — 

 injured or otherwise inihealthy ; in tiiose cases where these birds were still in 

 flocks such supposition can of course not be made, and we must conclude that 

 they are birds i'rom the most northern haunts of the species, which is proved in 

 the case of Mntdcilbi Jlaea thunber(ji, which j)asses through after the passage of 

 its allies, and which goes farthest north. Jloreover I agree with Count Zedlitz 

 (hat some of these do not breed in a given year, though for reasons unknown 

 to us. 



A iiuestion which has often been discussed is : whether birds migrate with 

 an extended front, or whether they follow strictly defined narrow routes. The 

 theory of the strictly defined routes is based upon the fact that in suitable localities, 

 such as river-valleys (which naturally afford much food), quantities of birds are 

 observed, and that an astonishing number of migrants are seen on certain points 

 of observation, especially small islands, like Heligoland, Fair Isle and others. 



The conclusion drawn from these observations is, however, erroneous. The 

 fact is, that birds must feed during migration, and therefore alight where food 

 is found, while they pass over arid or otherwise unsuitable land, and therefore 

 are not observed ; and the reason why Ileligolanil and Fair Isle and other localities 

 are apparently so miraculously favoured by bird migration is, that they afford 

 rest, food, and shelter in a wide area of uninviting sea, an<l that keen and able 

 observers have made them their observatories I 



The various migration routes, as depicted by Palmen and otiiers, are based 

 on meagre records, and most of them were api)arently drawn because tlie map 

 suggested them ; tlins migration routes I'ross the Mediterranean where it is 

 narrowest: west of Gibraltar; from Valencia to West Algeria (leaving Alger 



