2 26 THE MIGRATION OF BIRDS. 



height during September and October, being much 

 less pronounced towards the extreme dates or 

 periods of its duration. From these facts, and 

 those in the preceding chapter, it will be seen that 

 Migration is more or less in progress during ten 

 months out of the twelve, only lapsing for a period 

 of two months in December and January. As we 

 have already seen, however, even these two months 

 are characterized by certain local movements, so 

 that really migratory birds, generally speaking, can 

 never be regarded as in a state of absolute rest. 

 For various reasons the Autum.n Migration of birds 

 is not so observable as that in spring, at least so 

 far as the actual time of departure is concerned. 

 Of course when birds are on direct passage the 

 movement is easily enough remarked along all the 

 recognized routes, species after species passing 

 certain stations with considerable punctuality. It 

 is, however, a very difficult thing to note the exact 

 period of departure from the place where the 

 migration of a species begins. Most birds moult 

 just previous to migrating in autumn, and at that 

 time they become very skulking in their habits, 

 lose their voice, and make a practice of slipping so 

 quietly away that they are not missed perhaps for 

 days after they have really taken their departure. 

 The only reliable sign therefore that migration is 

 in progress is that of individuals actually en route 

 at a distance more or less remote from their starting- 

 place. The great wave of Autumn migrants begins 

 to flow south in July, gradually increasing in 



