( 1S4 ) 



wcro known as the homo of this interesting pijreon : Albeniiiile, Duncan, Charles, 

 Hood, rhathaiu, liidefatijralilc, James, Tower, Biiulloe. Banr adds (Am. J\V^^ 1897, 

 p. TS4) Jervis Island. We have received it from most of these ])laces, and in 

 addition from Abingdon, Narborough, and Gardner Islands. Specimens from all 

 these islands do not differ, as far as we can see, but we have received a large series 

 from Weumaii and Cali)eiiper, and fiud them to differ materially in size. 



Two femnles from Hood Island are semi-albinistic. Their tails are for the 

 greater part of their length light grey, while one of them has also the primaries 

 white with brown tips. 



2. Nesopelia galapagoensis exsul siibsp. nov. 



Differs from Nesopelia galapagoemis galapugoensis in being larger. While the 

 wiug of the latter varies in the male from 130 to 140 mm., it measures in our new 

 subsj)ecies 142 to 148 mm. As we have measured twenty from ('uljjepper and three 

 from AVenmau, we cannot be mistaken. Also i\\e^ females, of which we have only 

 one from Culpepper and two from Weumau, have the wing longer iha,n J'enudes from 

 the southern and central islands of the group — i.e., 130 to 135 — while in the latter 

 it varies between 120 and 120 mm. The tail is also about 5 to 8 mm. longer in the 

 birds from ('nlj)epjier and AVenman Islands. 



" The iris is brown, the bill lilack, the feet pinkish red, skin round the eye 

 indigo-blue." 



All the birds from Culpepper and Weunian Islands were collected in July. 



The femides of both forms differ from the males in their much smaller size, 

 duller upper surface, and more whitish wing-coverts, while the sides of the neck are 

 often as glossy as in the male. Immature birds are below brownish, with wliitish 

 fringes to the feathers, the sides of the neck are not glossy, and the lesser upper 

 wing-coverts have rufous edges. 



]\Ii'. Harris writes : " The e.xtreme scarcity of Nesopelia on Alliemarle, Charles, 

 and Chatham Islands is noticeable, as these islands are infested with wild house-cats, 

 and these pigeons, being principally ground-birds, are easily caught by the cats." 



Genus CRECISCUS Cab. 



CreciiCtis, Caliauis in .lniirn.f. Oni. 1856, p. 428. 



New World from temperate North America to Peru and Chili. If the genus 

 Cfeciseus is separated from Forzana at all, then the (Jalaj)agos rails belong 

 doubtless to Crecisc'ts. 



1. Creciscus spilonotus ((Jonld). 



Ziqmriua sj/iloiiola, J. Gould in Darioins Voij. Beagle, III. Birds, p. 132, PI XLIX. (1841). 

 Porzumi sjtilojiola, Salvin in Trims. Zoal. Sue. Land. IX. p. .500 (not v. X.) (partim !) ; Ridgway in 



Pror. U.S. Nat. .l/«.v. XIX. p. G18 (partim?). 

 Porzutia gtila/iaiioeiixin, Sharpe, Qil. H. Brit. .]his. XXHI. p. 113 ; Ridgway in I'mr. cr..S. .\at. .!/«». 



XIX. p. Gib. 



Dr. Sliai'pc was the first author who recognised that there were two forms of 

 rails in the Galajjagos Islands, one, a larger and more powerful one, discovered by 

 the collectors of the " Beagle" and apparently not found since, on James Island,* the 



* Darwin, Juuntal of Jlfseurchea into the Nat. llist. and Oitd. of countr. xis. ifttr. vmj. round tin- WorUl. 

 p. 402 (e<l. 1S90) : " So iliuap was tlic groiiiul that theie weic large beds of a coarec cj'perus, in which great 

 ntiniburs of a very bmall wutci-ruii lived uud bred," 



