JAN., I902. [ THE CONDOR 17 The composition is externally of twi?s, rootlets, moss and coarse grass, while the lining generally consists of fine root- lets. Occasionally, however, a nest is found thickly lined with horse and cow- hair. The number of eggs in a complete set varies from three to five, the latter number being very rare, while three is nearly as often found as four. In color the eggs have a bluish-green ground, rather sparsely spotted and dotted with colors varying from lavender to green- isb brown and almost black. A few specimens rather closely resemble some eggs of erythronzelas, but the vast major- ityhave a much deeper blue ground color and the markings have a less red- dish tinge. The markings generally tend to the fonnation.ofa ring around the larger end, but the eggs are always more or less marked over their entire surface. The variation in size is considerable, the largest in our collecting measuring .96x.67 inches, while the smallest is .86x-
67 inches. The average of twelve eggs
taken at randomsh6wsa measurement of.925x.654 inches. Of course none of the eggs mentioned are in any way ab- normal. These birds are seldom very solicitous about their nest and eggs. The female sits closely until forced to leave the nest, then flies to a short dis- ance and soon brings the male by her chipping, and soft purring notes. Both then come back and watch the intruder at a short distance, with but little dis- play of anxiety. Eastern Limit of Lawrence Goldfinch. In his 'Birds of North and Middle America,' Dr. Ridgway gives the east- ern limit of the range of Lawrence goldfinch ( flstra.?alDzus lawrence/) as Fort Whipple and P/hal County, Ari- zona. I have two specimens (?3 and ??) that l shot January 20, I876 near Fort Bayard, Grant County, New Mexico, about five miles east of the continental divide. F. STEPHENS. San Dt'e?o, Cal. Correction of Errors in Identification, I have been responsible for the fol- lowing erroneous records: (i) 7'rin.?a baird/ Gr?t?Lr,, Auk XV, April ?898, ?26. The specimen re- ferred to, taken at Sitka, later proved to be an'immature spotted sandpiper (Actills macMaria) (2) ?trdus aonalaschk?e auduboni GRI?rNE?r, Bds. Pac. Slope Los Ang. Co., March x898, 5I. The specimen proves to be f?.ylocichla a. aonalaschk?e, and not either ?. a. auduboni or ? a. sequoiensis. (3) Otocoris alpestris leucol?ema B?R- now, Condor I[I, Nov. ?9o?, x67. I was to blame for this naming. The specimens were since compared by W. K. Fisher at Washington, and pro- nounced to be Otocoris alpestris merrilli. These are all the mistakes in identifi- cation that I am so far aware of having made in any published writings. If others come to light, I propose to an- nounce them at once, so as to avoid further danger of such erroneous re- cords being quoted. I would urge that other writers do the same. For mistakes are bound to be made at some time or ancther, and uncorrected errors of this kind have caused much trouble in the past. The sooner they are made right, the better. In the study of distribution and its modifying influences, a few mistaken records may cause confusion, and per- haps prevent correct deductions. [ do not include here changes in names, due to shifting nomenclature, or to separation of newly-recognized geo- graphical forms. For the compiler will readily recognize the form meant usu- ally by the locality. But out and out blunders like the above, where the spotted sandpiper was recorded as bairdi, could not be judged as such, un- less admitted by the author, Such blunders we know to have been often made in breeding records; for instance, the "black swift" nesting at Seattle, and the "evening grosbeak" nesting in olo County. It should be the duty of those who know of such errors to point them out assuch, as soon as discovered. JOSEPH GRINNELL.
