320 XOVITATES ZOOLOGICAE XXV. 1918. 



Dutch New Guinea, but the bill is much shorter even than in females of the 

 latter, and the knob is smaller. Moreover, the underside is much darker. The 

 breast and abdomen are more greyish brown, of about the same colour as in the 

 much larger subtuberosus from Fergiisson and Goodenough Islands. Even if the 

 three specimens should be females (which we have no reason to suppose), they 

 would be of a quite distinct new subspecies. The bills measure 35' 6-38, the 

 wings 142-143 mm. 



Type: ^ ad. Sudest Island, 6. v. 1916. Eichhorn coll. No. 7411 of the 

 A. S. Meek collections. 



" Iris dark reddish brown. Bill black. Feet slate-colour." 

 Mr. Ogilvie-Grant, Ibis, Jubilee Suppl. ii. p. 78 (1915), expresses his opinion 

 that two of the subspecies wliich we had recognized in 1913, viz. brevipennis and 

 aruensis, are not recognizable. His view about brevipennis is not correct and 

 his own measurements rather support our separation of this form, but he only 

 measured four males of P. novaeguineae novaeguineue. In our series the wings 

 of males run up to 166 and 167 mm., a length wliich is never approached in 

 brevipennis. With regard to aruensis we expressed ourselves {I.e. p. 514) rather 

 hesitatingly, only saying that we were inclined to separate it, but we now be- 

 lieve that it should be separated, though it is close to P. n. novaeguineae. Ogilvie- 

 Grant says : " The difference in the colour of the head used as a character is 

 non-existent." We mentioned that the heads of aruensis " are after all gener- 

 ally much paler," and that is a fact, though sometimes no difference is visible. 

 But there is another character which requires attention : the knob in aruensis 

 rises more gradually, so that it becomes longer and the beak in front of the knob 

 higher, whereas in novaeguineae it rises more abruptly so that there is a much 

 sharper declivity in front of it. This is very obvious in the series, though a few 

 specimens intergrade in this respect and are not distinctly separable on this 

 character alone. 



Erythrura tricliroa goodfellowi O.-Grant. 



Erythrura trichroa goodfellowi Ogilvie-Grant, Bull. B.O. Chib, xxix. p. 29 (1911^ — Moroka Moun- 

 tains) ; R. & H., Nov. Zool. 1915. p. 36. 



One male from Mt. Riu or Rattlesnake, 25. ui. 1916, No. 7316, agrees with 

 our series of about 20 adults from British New Guinea and Vulcan and Dampier 

 Islands. " Iris dark brown with ashy blue outer circle. Bill black. Feet Ught 

 transparent brown." 



Alcedo atthis pelagica Stres. 



[Oracula atthis Linnaeus, Syst. Nal. ed. x. I. p. 109 (1758 — Eg\pt. Ex Hasselquist, iter. Cf. Laub- 



mann, Verli. Orn. Ges. Bayern xii. 4, p. 238. 1916).] 

 Alcedo ispida pelagim Stresemann, Nov. Zool. 1013. p. 310 (Eastern Now Guinea, D'Entrecasteaux 



and LoiiLsiade Islands). 



Wings 68-70'5, in one male, however, 73 mm. This form was separated 

 by Stresemann on account of a shorter average of the wing, which he measures 

 66-72, but he had overlooked the <J A\ith a wing of 73 mm. As the nearest 

 form, A. atthis hispidoides Less., has the wing 68'5-75 mm.; our pelagica has 

 it only about 2-2' 5 mm. shorter. Twenty-one pelagica were compared with 57 

 hispidoides. 



It is of course very important to register even such small diSerenees in size, 



