ORNITHOLOGICAL EXPLORATIONS. 51 



Other misstatements as to the color of the naked and soft parts have 

 originated in describing them from dried skins, or at least from speci- 

 mens not absolutely fresh. This is evident, for instance, where the bill 

 and feet of the young birds are described as black (Orn. Centr., 1882, 

 p. 29 : " Beine immer noch schwarz "), and th*e rosette of the adults as 

 " pjEonia-red." 



Another mistake is probably due to lack of material, the stage in 

 question not being represented in the " splendid " series. I allude here 

 to his statement (Orn. Centrbl., 1882, p. 30) that when the horny orlet 

 falls off it is replaced by MacTc feathers. The fact is that the feathers 

 which are growing out when the orlet drops off are ivhite, as my speci- 

 mens conclusively prove ; this part drops off before the white face- 

 mask has changed to black ; when that change takes place then the 

 white feathers on the ''membranous orlet" are first exchanged for 

 black ones. 



The series of plain and colored drawings which accompany this ac- 

 count were carefully prepared on the island from specimens perfectly 

 fresh and still in the flesh ; they were not merely sketches, hut finished 

 on the spot, and have not been touched since. I claim for them the 

 merit of being true in outline and color. They illustrate most of the 

 different stages of development and transformation, and, in conse- 

 quence, are frequently referred to in the following, as they certainly 

 convey a better idea of the specimens than the most explicit descrip- 

 tion ; besides, the descriptions have been given above under the re- 

 marks following the '' list of specimens collected," p. 44. 



The color and shape of the bill of the downy young (j)!. ii, fig. 1) is 

 exactly that of the more developed young represented in the same plate, 

 fig. 2. Passing through the stage of fig. 3, the bill grows gradually 

 during the winter, simultaneously reddening in the anterior part. Birds 

 hatched very early will have attained a development corresponding to 

 fig. 4 at the middle of February, while late birds at the same time will 

 not have advanced farther than the stage represented by fig. 3. The 

 next step is the swelling and growth and subsequent hardening of the 

 basal parts accompanied by their changeof color into an olivaceous green. 

 At the same time the semilunar vertical grooves on the anterior part 

 of the upper mandible commence to be visible. It seems as if they are 

 formed by a constant flaking oft", as the superficial corneous layer of 

 the grooves is always more or less lacerated, detached, and in course 

 of scaling, causing the whitish coloration of the grooves. When this 



