240 



AVIFAUNA OF LAYSAN, ETC. 



accepting the name of King (" B alius ecaudotus") for the little Hail from Hawaii with 

 unspotted notseum, in opposition to ■« Ballus sandwchensis" with a spotted upper surface. 



From all we know at present of the specimens of Sandwich Island Rails (namely, one in 

 Leyden with a spotted hack, and five with a uniform upper surface, of which three are now 

 in England and two in Honolulu), I must, together with Wilson, Hartlaub, and Sharpe, 

 recognize two distinct species. For one of these (the Leyden specimen) I accept the name 

 of Gmelin: Ballus sandwichensis, based upon Latham's "Sandwich Rail." It is distinctly 

 said that " the feathers on the upper parts are darkest in the middle " ; and in the Latin 

 diagnosis (Index Ornith. /. c.) f "Ballus pallide ferrugineus, supra maculis obscurkr 

 Altogether Latham's description suits the Leyden bird very well, in fact about as well as any 

 of Latham's descriptions suit a bird known to us. This has already been recognized 

 by Dr. Ilartlauh. It is also known from the sale-catalogue of the Bullock collection that 

 Temminck bought a Sandwich Island Rail from that collection in 1810 (Hartlaub, I.e.). 

 Although the JPennula now in Leyden is not labelled to that effect (Finsch, I. r.), we can 

 hardly doubt that it is the same specimen ; and it is— notwithstanding the doubts of Dr. Finsch 

 —quite possible that this is one of the birds described by Latham, many of which had passed 

 into the Bullock collection. Finsch, it is true, has cleverly proved that the Leyden Rail is 

 not necessarily and beyond doubt the type of Latham's Sandwich or Dusky Rail; hut after 

 all there is a possibility that it is so. It is certain that only some, not all, of Cook's 

 specimens were merely dried and not skinned, as 1 have been told by Mr. Robinson of 

 Liverpool. But, accepting Dr. Finsch's assertion that his bird is not a " type," I fail entirely 

 to see the reason for giving it a new name. Dr. Finsch does not at all show that his bird 

 disagrees with Latham's diagnosis. It is true that Mr. Wilson {I. c.) says that " the specimen 

 does not correspond with the Sandwich Rail of Latham"; hut I agree with Dr. Hartlaub 

 that it does. There is therefore no reason for giving a new name to the Leyden Museum 

 bird, because nobody will follow Dr. Finsch in renaming a specimen only because it is not 

 " the type "—for this is exactly Dr. Finsch's case. 



The best proof, however, for the identity of " Hollas sandwichensis " with " Fennula 

 wllsoni " is the drawing of Ellis, now published by Mr. Scott Wilson, which agrees exactly 

 with Latham's description. Comparing this with Mr. Frohawk's and Mr. Keulemans's 

 figures taken from the Leyden specimen, one must notice the most striking resemblance, the 

 diilerenccs being nothing but a paler colour of the old drawing and a want of artistic skill ; 

 but then it must be remembered that the date of these plates is about 120 years apart, and 

 that our present men are more or less accomplished ornithological artists, while Ellis was 

 certainly (cf. plate) not an artist in bird-drawing. I can easily show plates done quite 

 recently from the same specimens hy different draughtsmen which differ as much and even 

 more than the plate of Ellis and those of Keulemans or Frohawk. 



If Latham's "Dusky Rail" was really from the Sandwich Islands, it was probably the 

 same as his "Sandwich Rail," as it cannot refer to Petmula millsi, having a streaked upper 

 surface. 



