58 THE BIRDS OF HELIGOLAND 



Thrushes, which in jiart swarm round the lantern and settle every- 

 where on the plain of the Highland, are principally captured — 

 sometimes in astonishing quantities. Thus, on the night of the 

 6th of November 1868, no less than 15,000 Larks were caught in 

 the space of about three hours. Unfortunately, the moon rose at 

 ten o'clock and put an end to the capture. No less than 3400 of 

 these birds, besides innumerable Starlings, several Snipe, and many 

 Blackbirds, were caught against the panes of the lighthouse lantern 

 alone. It is therefore impossible to form even a remote estimate of 

 the figure to which the spoil would have mounted if there had been, 

 for the whole of this remarkable migration night, what is known as 

 a ' dark moon.' 



The appearance of a single star, or a fragment of blue skj^ 

 through the dense and uniform blackness of the night, or the faint 

 gleam on the horizon which announces the rising moon, is followed, 

 as on the date above mentioned (6th of November), by the imme- 

 diate disappearance of the crowds of wanderers with whose manifold 

 voices the air had been fiUod but a few moments before : all having 

 risen to heights to which it is impossible for the eye to follow : nor 

 can even a single call-note be any longer heard from afar. Neverthe- 

 less, there is no interruption to the stream of migration itself ; for 

 if, half an hour, or an hour or two afterwards, dense and uniform 

 darkness should 'once more envelope the sky, the birds may teem 

 again on aU sides, and the capture proceed anew. 



What has been said above illustrates in the clearest manner how 

 an apparently slight change in atmospheric conditions at once 

 influences the height of the migration flight, and how Uttle is 

 required either to bring this flight within our powers of observation 

 or to withdraw it from the perception of our senses. Here I must 

 once more repeat my opinion as to the comparatively insignificant 

 value of recording data of the occurrences of birds at definite 

 points. To control a district about four miles in diameter, and pre- 

 senting the varied features of wood, moor, corn-fields, meadows, and 

 water, is in itself almost an impossibility. How would one be able 

 to determine day by day what difl'erent species have occurred in 

 such ditierent localities I The case of Heligoland is, of course, 

 diiferent, for here we may say without hesitation that literally not 

 a single bird escapes observation. Notwithstanding, the results of 

 notes of this nature can never amoimt to anything more than a list 

 of disturbances and interruptions of the main migration movement 

 at the particular place of observation, although the investigation of 

 the causes of such disturbances is undeniably a study of the highest 

 interest. Any other results of such records, even if these have been 

 kept for a long series of years in a favourable locality and with 



