( 227 ) 



3. Baza subcristata reinwardti CMiill. & Schleg.). 

 Two mali:s, Kayoli, Bam (cf. <uifeii, \). 20). 



4. Ninox hantu (Wall.). 

 S ?. Doherty collection, Kayoli, Blarcli 1807. 



The relations of N. kcmtii require consideration, but I am not able at present to 

 work them ont. 



5. Pisorhina magica (S. Miill.). 



c? jnn., Kayeli, Doherty collection ; cJ ad. Monnt Mada, Dumas collection. 



With this bird we have again reached a most vexed and difficnlt crnx of 

 systematic ornithology, which has been treated differently by almost every author. 

 Sharpe {Cat. B. Brit. Mus. II. pp. 69 — 7.5) has separated Scops niagkus magicus 

 (Ceram, Amboina), S. magicus leucospiliis (Batjan, Gilolo), S. magicus boiiruensis 

 (Bonrn), S. 7nagicus morotensis (Morty), and others, which I will not mention, as they 

 are probably quite distinct specifically. Schlegelunited all these forms. Salvador! 

 {Orn. Pap. I.) recognised S. magicus (Amboina, Ceram (?), Aru), S. leucospilus 

 (Batjan, Halmahera, and Burn), and iS'. morotensis (Morty, Ternate). 



Meyer and Wiglesworth {B. Celebes I.) mention magicus (Amboina, Ceram), 

 leucospilus (Batjan, Halmahera, Ternate, Burn), and morote?isis (Morty, Ternate), 

 all as subspecies of manadensis. Quite recently we have a most jjainstaking article 

 on Scops magicus and its allies by Dr. Finsch. He {Nat. Lej/den Mus. xx. p. 103) 

 unites magicus, bouruensis, leucospilus, and morotensis, and gives the following 

 distribution : Amboina, Ceram, Burn, North Celebes, Batjan, Ternate, Halmahera, 

 Morotai, Aru, Sumbawa. There are fifty-four examples in Leyden, in different 

 reddish and brown phases. Sharpe, Schlegel, and Finsch have in any case shown 

 laudable consistency ; the former in sjditting them up to the ,bitter end, the latter in 

 uniting them all. Salvadori's (and Me3'er and Wiglesworth's, after Salvadori's) view 

 can hardly be correct, as ho evidently went by the different phases of coloration and 

 by the markings, which are very variable, and constructs a distribution which 

 is difficult to believe in, allowing leucospilus to spread over the Northern and 

 Southern Moluccas (Buru), and morotensis over Morty and Ternate, but not over 

 Halmahera, where, between the two strongholds of morotensis, the allied leucospilus 

 is to occur I In view of the close relationship of these forms, which must be 

 admitted by every one, such a distribution is not probable ; it is possible if 

 these forms are all "good species," while it can not be accepted if Meyer and 

 Wiglesworth's view of their subspecific value is taken. 



As far as my conception of these owls goes, the case is as follows : — 



The specimens from the Southern Moluccas — namely, from Amboina, Ceram, 

 and Buru — are absolutely identical in a specific as well as subspecific sense. Those 

 from the Northern Moluccas — namely, Batjan, Halmahera, and probably also Morty 

 — are identical inter se ; but it seems to me that the northern form might be separable 

 subspecifically as a smaller race from typical magicus from the Sontli Jloluccas. 



For the latter {magicus) I obtain the following measures of the wing : 107 

 (Sharpe), 177 (Sharpe), 185, 186, 188, 190 mm. ; for the smaller northern form : 

 157 (Sharpe), 172, 175, 178. Finsch gives for the latter 102—191, for the former 

 173 — 192 mm. My measurements and Dr. Sharjie's show a striking difference 

 between the northern and southern forms, while Finsch's exhibit only a very small 



