GANNET. 77 



white, wedge shaped on the back, scapulars, wing-coverts, head, and hind-neck, 

 and edged with white at the tips of the primary -quills ; under-surface for the most 

 part white, with some brown feathers on the throat, fore-neck, and thighs. 



Nestling. Covered with white down. 



Nest. Composed on the ground, shaped like a flat cone from 4 to 6 inches high. 

 Egg cavity from 6 to 8 inches by one to two deep. 



Eggs. Clutch, one ; lime covered ; axis 78 to 82 mm., diameter 49-50. 



Breeding-season. October to January. 



Distribution and forms. Extra-tropical Australia and New Zealand. The 

 Australian birds have the colouring of the head and neck of a much lighter shade 

 than the Neozelanic birds and consequently we must admit two subspecies : S. 8. 

 senator Gray, from Tasmanian seas, and 8. s. rex subsp. nov. from New Zealand. 

 This alteration is necessary through the fact that Gray quoted S. serrator as a 

 synonym of S. australis Gould, and when the latter name proved to be preoccupied 

 the former comes into service in the same connection and with the same type 

 locality. 



ORDER LARI. 



This order is a complex of sea and wading birds of different aspects and not 

 easily circumscribed owing to their development in different directions, some having 

 become almost entirely seabirds, others partly seabirds, and others purely land 

 birds. 



We subdivide the order into seven suborders, of which three have no repre- 

 sentatives in the Australian fauna These are the Alciformes, Colymbiformes and 

 Chionidiformes, the former two being confined to the Arctic and North Temperate 

 zones, the latter including two families one. the Chionididae, of sub-antarctic distri- 

 bution ; the other, the Dromadidae, of tropical range, and probably with little real 

 affinity. As to the Alciformes and Colymbiformes, both of which are restricted to 

 the Arctic and North Temperate Regions, we may quote Ridgway, who concluded : 

 " Undoubtedly these are closely related ; but the evidence seems to point to a closer 

 relationship of the former to the Gulls than to the latter." We suggest that their 

 development has been from a Larine ancestor on somewhat similar lines, the differ- 

 ences observed being such as would be satisfied by such an hypothesis. Probably 

 the Skuas are almost as far away, but their line of descent has somewhat approxi- 

 mated to that of their Larine relations, with which they are associated. The litera- 

 ture of the anatomy and osteology of the Chionidiformes is vast and the results appear 

 to be that both are somewhat generalised forms agreeing in items with the suggested 

 ancestral Limicoline form. They are as peculiar in anatomical details as they are 

 superficially, and the external items appear just as suggestive as any internal ones. 



The whole group shows so much variation in the deeper-seated features that 

 no diagnosis can be drawn up covering the series, every character showing exceptions. 

 Thus the basic position has been denied the Larine forms, as basipterygoid processes 

 are absent in the adult, though present in the juvenile, while they persist in the adult 

 in the Charadriine group ; such an argument seems futile in view of its variation in 

 this order. 



SUBORDER PHAETHONTIFORMES. 



This suborder consists of seabirds with straight heavy bills, long wings, wedge- 

 shaped tails with two very elongated and attenuated central feathers and short legs 

 with the four toes webbed, the middle toe longest, and the hind-toe small. The 

 suborder covers the only family of few species, formerly referred to a single genus, 

 but now this is acknowledged as incorrect. These are all tropical in habitat, ranging 



