A MONTH ON THE EDDYSTONE 307 



On each occasion when a number of birds of any 

 species was killed at the lantern, it was interesting to 

 note how considerably they varied in size, and some, 

 though to a less degree, in colour. The Skylarks, 76 

 in number, obtained during the great movement of 12th 

 to 13th October, showed the remarkable range in wing- 

 measurement of from 4.70 in. to 3.85 in. ; the Starlings, 

 obtained on the same date and 53 in number, ranged 

 from 5.38 in. to 4.85 in. ; and the Meadow-Pipits from 

 T^.-^j in. to 2.91 in. The Skylarks and Meadow-Pipits 

 exhibited some variation in colour, difficult to describe 

 in words, but quite manifest to the observer. It is 

 possible that more than one race of the two last-named 

 species was represented during the movement, or it may 

 be, in the case of all three species, that the variations 

 in size, etc., were due, in a greater or lesser degree, to 

 age or sex, or both in combination. 



Wing-measurements are valuable as an indication of 

 the range of variation within species, but speculations 

 based upon ordinary material are apt to be extremely 

 misleading. Here, again, sex and age, singly or in 

 combination, may, and do, account for much of the 

 variation to be found, and yet how insignificant are the 

 data in our possession which afford these essential 

 particulars ! 



As bearing directly upon these remarks, I will 

 instance a few cases that came under my notice at the 

 Eddystone. In addition to those of the Skylarks and 

 the Meadow-Pipits (which showed a very considerable 

 variation in size and certain diversities of plumage 

 inte7^ se, though all were obtained during a single move- 

 ment), the Starlings killed on the night of 12th to 13th 

 October were all of one race, namely, the purple-headed 



