PLUMAGE i6 S 



were collected by myself, but no attempt was made to pick out examples of any particular 

 plumage phase — the birds were killed quite at random. 



The thirty-seven specimens were all collected between November and April and none 

 shows any sign of moult. Since the skuas leave the Falklands in April it follows that 

 the moult must take place during the southern winter when the birds are absent from 

 their breeding haunts, thus affording a contrast to the habit of McCormick's Skua which 

 moults after the breeding season (Wilson, 1907, p. 74). 



The birds from the Falkland Islands undoubtedly belong to one race, and even a 

 superficial examination shows that it is very variable ; dark, light and intermediate forms 

 are all present. If the skins are arranged in order of size (as judged by wing length) it 

 will be found that there is no correlation between colour and measurement. 



As has been said on p. 164 the differences between the subspecies of Brown Skua are 

 stated to be those of plumage and size, but in every subspecific group there is con- 

 siderable individual variation, and so far as colour is concerned the specimens of any 

 race may be matched by skins from the Falklands. There are, for example, two or three 

 skins which have the dusky plumage supposed to distinguish Mathew's lonnbergi and, 

 contrasting with them, are very light birds, plentifully splashed on the mantle with pale 

 marks, which are identical with South Shetland specimens of clarkei. In view of so 

 much agreement I do not consider that plumage differences can be regarded as valid 

 subspecific (racial) characters. 



MEASUREMENTS 



The subject of differences based on dimensions may now be considered. The four 

 standard measurements are those used by Lowe and Kinnear and are the following: 



(1) Length of wing from the carpal joint to the tip of the longest primary, the manus 

 having been straightened out as much as possible. This measurement must be considered 

 as only approximately correct; complete accuracy is impossible since the amount of 

 bending of the dried manus varies from one specimen to another and it is never possible 

 completely to straighten it out. 



(2) Length of exposed culmen. 



(3) Depth of bill at base of exposed cidmen. 



(4) Length of tarsus. 



Since it is obvious that the greater the variability of a character the less is its value in 

 the separation of subspecies, I have attempted to determine the range of variation of 

 the four measurements in each of my geographical groups. The measurements are shown 

 in Table I and Fig. 1. Table II gives for each group firstly the average, minimum and 

 maximum measurements, and secondly the divergence of the minimum and maximum 

 from the mean, expressed as percentages of the latter, and range of variation also ex- 

 pressed as the percentage of the mean. 



It will be observed that the lengths of the wing and tarsus are the least variable of 

 the measurements, and that the depth of the bill shows greater range than the length. 



