3i8 THE NESTS AND EGGS OF 



little doubt that it pairs for life, being observed in pairs 

 a month or more before the eggs were laid. The birds 

 arrived at their breeding haunts towards the end of 

 November, and as usual kept pretty close during the 

 day, going out to sea at dusk to feed. The favourite sites 

 for the colonies were rock and stone-strewn slopes on 

 the sides and summits of high hills, or amongst the 

 shattered rocks just above high-water mark on the beach. 

 No nest is made, the ^g^ being laid on the bare ground 

 in hollows under stones or beneath masses of rock. 

 When the colonies were visited at night, the birds were 

 found to be noisy, and flying to and from their holes, 

 uttering their notes even vv^hilst on their eggs, but when 

 alarmed they became silent, as if anxious not to betray 

 their whereabouts. 



Range of egg colouration and measurement : 

 Wilson's Petrel only produces a single ^'g% for a sitting. 

 This is white in ground colour, generally with a more or 

 less distinct zone of dust-like specks of reddish-brown 

 round the larger end. Average measurement, \"^ inch 

 in length, by '9 inch in breadth. Incubation is performed 

 by both sexes, but the duration of the period is un- 

 determined. 



Diagnostic characters : The eggs of this Petrel 

 cannot be distinguished from those of the Fork-tailed 

 Petrel, but as the latter only breeds in the Northern 

 Hemisphere, the locality is sufficient to separate them. 

 Whether the eggs can be distinguished from those of 

 other Petrels breeding in the Southern Hemisphere, lack 

 of material prevents me from stating. 



