Two eggs of G. newtoni, from St. Croix, are of a more rounded-oval shape, and measure .69 by .45 and .65 by .44 of an inch. They have a dull white ground, but this is so uniformly and generally covered with confluent reddish-brown markings as to be nowhere very distinct.

The St. Croix species is called the Sugar-Bird in that island, from its habit of entering the curing-houses, through the barred windows, probably attracted thither by the swarms of flies. It is a very familiar species, haunting gardens, and often entering houses, and never manifesting any alarm. It keeps in pairs, and breeds from March to August. Mr. Newton states that it builds a domed and often pensile nest, with a small porch, or pent-house roof, over the entrance, generally at the extremity of a leafy bough. The nest is generally very untidy on the outside, and is composed of coarse grass and cotton, with feathers on the inside. It deposits its eggs before the completion of the nest, “rather to the discomfiture of the oölogist, who delays inserting his finger into the structure while he sees one or both of the birds busy with a tuft of grass or cotton in their bills, until at last he finds their eggs already hatched.” Mr. Newton observed one instance in which two broods were reared in the same nest, with only an interval of ten days between the time the young left it and the laying of an egg.

Family TANAGRIDÆ.—The Tanagers.

Char. Primaries nine. Bill usually conical, sometimes depressed or attenuated, usually more or less triangular at base, and with the cutting edges not much inflected; sometimes toothed or notched. Legs short; claws curved; colors usually brilliant.

We confess our entire inability to present a diagnosis that shall define and separate satisfactorily by external characters the closely allied families of Cærebidæ, Sylvicolidæ, Tanagridæ, and Fringillidæ, agreeing as they do in the main in every respect. The only attempt at distinction is based upon the shape of the bill, and this in what are generally called Tanagridæ presents every variety of shape, from the attenuation seen in Dendroica to the stoutest form of the Fringillidæ. The Cærebidæ have peculiarities of the tongue, not appreciable, however, in the skin. In view, therefore, of the difficulty in question, we shall copy the conventional names and unsatisfactory definitions of other authors, in our inability to present a satisfactory arrangement of our own.

Carus and Gerstæcker in Handbuch der Zoologie, I, 277, adopt a classification of the Oscines based on the palatine bones in which Fringillidæ and Tanagridæ are distinguished from the Sylvicolidæ as follows:—

Suborder OSCINES, Sundevall. Of the ten primaries, the first is short, rudimentary, or wanting; the number of secondaries is rarely more than nine. Tarsus entirely booted, or else with an undivided plate on the sides. Lower trachea completely formed by the help of the trachea and bronchiæ; generally with four pairs of muscles, distributed before and behind.

Group I. Spizognathæ. Outer lamella of the palatine bone developed in a vertical plane, with the hinder border more or less emarginated; the anterior palatine process broad, and united by a truncated border to the high and broad upper mandible.

Family 1. Ploceidæ. Ten primaries.