Nos. 2 & 3.] THE AUSTRAL AVIAN RECORD 33 



We have instanced this case in detail, but do not propose to 

 go into such in connection with the succeeding orders. 



In the Euornithes we begin with the Procellarii or Tubinares 

 and admit only one suborder, Procellariiformes, divisible into 

 four famihes, Thalassidromidse, Procellariidge, Pelecanoididge 

 and Diomedeidge. It would be a good system to amalgamate 

 the three first named into a superfamily, Procellarioidea, and 

 contrast the latter as a superfamily, Diomedeoidea. The 

 latter better represents the value of the divisions, as we 

 cannot yet give them subordinal distinction. 



As a distinct Order we must separate the Fregati, and 

 we place these next, as they show structural characters 

 recalling those of the previous family, and, moreover, 

 these are confirmed by internal features, and all investigators 

 are agreed as to the anomalous position these held when 

 included in the Steganopodes. We restrict the latter to those 

 members left when the Fregati and Phaethonti are removed ; 

 and, admitting that only one suborder is necessary, would 

 still differentiate the groups as being above family rank, and 

 therefore recognise three superfamilies, Phalacrocoracoidea, 

 Pelecanoidea and Suloidea. Two families are included in the 

 first named, the Phalacrocoracidse and Anhingidse, while the 

 others cover but one each. 



The Order now following, the Lari or Limicolse, is an amal- 

 gamation of several suborders, Phaethontiformes, Lariformes, 

 Alciformes, Colymbiformes, Chionidiformes, Charadriiformes 

 and Otidiformes. The first named is, according to our judg- 

 ment, more closely related to the Lariformes and seems a 

 derivative from the ancestors of the Sternidse series. On 

 account of its " steganopod " character we place it at the foot 

 of the Lariform group, succeeding with the Lariformes s. str.^ 

 with their usual four famihes, Sternidge, Rynchopidse, Laridse 

 and Stercorariidse. The difficulty of arrangement is felt 

 throughout the whole of this series as we trace the evolution 

 of many groups in different directions ; as here we can foUow 

 the Sternidse into the Rynchopidse, which must be recognised 

 as highly specialised, yet from their pecuhar distribution, a 

 very old form. Again, from their structure, the Stercorariidse 



