( 391 ) 



the comparison of a series of fully adult Duncan males would prove them to belong 

 to a lara'er snhspecies. 



The Barriiigton specimens, which we also mentioned under this head, although 

 we found them to be ratlier small, unfortunately ou re-examination evidently belong 

 to another form, which we ct)nsider will prove to be a new subspecies of G. fortis ; 

 but at present our material is too scanty to warrant our applying a new name to it. 

 The very young bird mentioned by us on p. 101 (under <!. fortis forti.s) belongs also 

 to this latter form. 



Thfi female from Indefatigable is in no way distinguishable from typical adult 

 females. 



Geopiza dnbia diihia was nesting on Chatham Island in March. The nests 

 were placed from 7 to 12 ft. above tlie ground, and were constructed of dry weed 

 and grass stems and moss, lined with finer grasses and moss, and mostly with 

 some cotton-balls on the outside. These eotton-))alls were noticed on nearly all 

 the Geospiza nests on Chatham Island. 



The nests sent show their shajie very well, the entrance being lateral, about 

 5 to 7i cm. wide, tlie eggs placed at tlie bottom, well downwards in the nest. 



The eggs are dull white, covered with pale rufous-brown patches and spots, 

 which are more frequent around the thick end. iVIeasurements : 20-8 x 15-8, 

 20-7 X 15 o, 2Mxir,-2, 20-4x10 mm. 



Geospiza dubia albemarlei llidgw. 



Nov. Zooi.. VI, 189;», p. \f>0.—G(oi>pizit alhrmaih'i Ridgw,, Dirdx North and MUhUc Aiucr. I, 1901, 

 p. 502 ; Geosjuzii firlix jilalj/rJiijHrJia Heller & Saodgrasa, C'omhir III., 1901, p. 75. 



"We liave received from both Beck and Green over 80 specimens of this form, 

 and on first comparing them we were inclined to think that Messrs. Heller and 

 Snodgrass were right in separating the .South Albemarle birds. However, after a 

 more careful comparison, we found not only both larger and smaller examples from 

 North and South Albemarle, but also a number of intermediate specimens of every 

 gradation, and we are therefore compelled to sink jilati/r/ii//ii//i'i as a synonym of our 

 Geospiza (labia alheiiiarlei. 



Mr. Beck found (/. dubia albemarlei nesting at Tagns Cove and C!ape Berkeley 

 at the beginning of April. He found the nests, though varying much in size, 

 generally larger than those of G. fortis, but otherwise they are similar. The two 

 nests taken were i)laced in bushes, 6 ft. from the ground, built of dry grass and 

 lined with finer grass, sometimes filso ornamented with some cotton-balls. 



Three clutches, two eggs each, are sent. The eggs of two of these clutches 

 have a decided, though very faiut, greenish tinge, while the third pair of eggs are 

 quite without a greenish tinge, being rather somewhat creamy white. 



The markings on some of these eggs are rather clear, and on one clutch the paler, 

 deeper-lying patches, and the darker, sharper surface-spots can easily be dis- 

 tinguished, while the others are more uniformly and somewhat duller marked. 

 Measurements: clutch 793— 20-1 x 15-4 and 20-3 x 15-4; clutch 791— 2M x l.'3-4 

 and 21-4 X 15-4; clutch 792-20x15-5 and 20x10 mm. The last clutch is the 

 one without a greenish tinge. 



