observed at tfi< j Kentish Knock Lightship, 131 



Office, wherein are shown the conditions prevailing over the 

 whole of the western half of Europe. 



The main weather-feature for investigation was naturally 

 that associated with the E. to W. movements. These 1 find 

 did not set in fully until a decided fall in temperature 

 took place in Western Central Europe, and this important 

 factor was the precursor of each of the pronounced move- 

 ments observed. Before such incentives to migration were 

 experienced, unusually high temperatures had prevailed, and 

 this was undoubtedly the reason why the movements prior to 

 the 8th of October had been of such a straggling and feeble 

 nature. These falls in temperature were not on all occasions 

 experienced on our shores, and this again demonstrates the 

 necessity for consulting the meteorological data at the place 

 where such movements have their source. 



During the great movements from E. to W. the direction 

 of the wind varied from S.E. to N.N.W., and possessed 

 no particular significance. The strongest wind prevailing 

 when I observed migration in progress was on the 13th of 

 October, when, with a westerly moderate gale blowing with 

 a velocity of from 34 to 40 miles an hour, Swallows were 

 proceeding in numbers to the S.S.W., and some House- 

 Martins to the N.W. — no other species being on the 

 move. Under like conditions on the 9th, Martius were 

 the only migrants observed, and were moving from N.W. 

 to S.E. 



The weather-conditions under which the other movements 

 were witnessed do not call for any special remarks, for my 

 experiences were similar to those at the Eddystone, and 

 supported the views already expressed in my previous 

 " study" on the bearing of meteorology on bird-migration. 



As at the Eddystone, whenever a number of individuals of 

 a species were obtained during any movement they shewed a 

 considerable range of variation in their wing-measurements, 

 bearing out fully what I have previously said on this subject 

 (see 'Ibis/ 1902, p. 267). As shewing how much individu- 

 ality may enter into these measurements, it is of interest to 



k2 



