the true relations of the members embraced under those heads. Enough has also been adduced to enable us to group many forms into families and somewhat more comprehensive groups, definable by osteological and other characters. Such are the Charadrimorphæ, Cecomorphæ, Alectoromorphæ, Pteroclomorphæ, Peristeromorphæ, Coracomorphæ, Cypselomorphæ, Celeomorphæ, Aëtomorphæ, and several others. But it is very doubtful whether the true clew to the affinities of the groups thus determined has been found in the relations of the vomer and contiguous bones. The families, too, have been probably, in a number of cases, especially for the passerine birds, too much circumscribed. The progress of systematic ornithology, however, has been so rapid within the last few years, that we may be allowed to hope that in a second edition of this work the means may be furnished for a strictly scientific classification and sequence of the families. (T. N. G.)
A primary division of recent birds may be made by separation of the (a) Ratitæ, or struthious birds and their allies,—in which the sternum has no keel, is developed from lateral paired centres of ossification, and in which there are numerous other structural peculiarities of high taxonomic import,—from the (b) Carinatæ, including all remaining birds of the present geologic epoch. Other primary divisions, such as that into Altrices and Præcoces of Bonaparte, or the corresponding yet somewhat modified and improved Psilopaedes and Ptilopaedes of Sundevall, are open to the serious objections that they ignore the profound distinctions between struthious and other birds, require too numerous exceptions, cannot be primarily determined by examination of adult specimens, and are based upon physiological considerations not necessarily co-ordinate with actual physical structure.
In the following scheme, without attempting to indicate positive taxonomic rank, and without committing myself finally, I present a number of higher groups into which Carinate birds may be divided, capable of approximately exact definition, and apparently of approximately equivalent taxonomic value. Points of the arrangement are freely drawn from the writings of various authors, as will be perceived by those competent to judge without special references. I am particularly indebted, however, to the late admirable and highly important work of Professor Sundevall,[4] from which very many characters are directly borrowed. The arrangement, in effect, is a modification of that adopted by me in the “Key to North American Birds,” upon considerations similar to those herewith implied. The main points of difference are non-recognition of three leading groups of aerial, terrestrial, and natatorial birds,—groups without morphological basis, resting simply upon teleological modification; a general depreciation of the taxonomic value of the several groups, conformably with the considerations presented in the preceding pages of this work; abolishing of the group Grallatores; and recognition of a primary group Sphenisci.[5]
A. PASSERES.[6] Hallux invariably present, completely incumbent, separately movable by specialization of the flexor hallucis longus, with enlarged base and its claw larger than that of the middle digit. Neither second nor fourth toe versatile; joints of toes always 2, 3, 4, 5, from first to fourth. Wing-coverts comparatively short and few; with the exception of the least coverts upon the plica alaris, arranged in only two series, the greater of which does not reach beyond the middle of the secondary remiges.[7] Rectrices twelve (with rare anomalous exceptions). Musical apparatus present in greater or less development and complexity. Palate ægithognathous. Sternum of one particular mould, single-notched. Carotid single (sinistra). Nature highly altricial and psilopædic.
a. Oscines.[8] Sides of the tarsus covered in most or all of their extent with two undivided horny plates meeting behind in a sharp ridge (except in Alaudidæ; one of the plates imperfectly divided in a few other forms). Musical apparatus highly developed, consisting of several distinct pairs of syringeal muscles. Primaries nine only, or ten with the first frequently spurious, rarely over two thirds the length of the longest, never equalling the longest.
b. Clamatores.[9] Sides of the tarsus covered with divided plates or scales variously arranged, its hinder edge blunt. Musical apparatus weak and imperfect, of few or incompletely distinguished syringeal muscles (as far as known). Primaries ten with rare exceptions, the first usually equalling or exceeding the rest.
B. PICARIÆ.[10] Hallux inconsiderable, weak or wanting, not always incumbent, not separately movable by distinction of a special muscle, its claw not longer than that of the middle toe unless of exceptional shape (e. g. Centropus). Second or fourth toe frequently versatile; third and fourth frequently with decreased number of joints. Wing-coverts for the most part larger and in more numerous series than in Passeres, the greater series reaching beyond the middle of the secondary quills (except in many Pici and some others). Rectrices commonly ten (eight to twelve). Primaries always ten, the first only exceptionally short (as in Pici). Musical apparatus wanting, or consisting of a muscular mass, or of not more than three pairs of syringeal muscles. Palate desmognathous or ægithognathous. Sternum of non-passerine character, its posterior border entire or doubly notched or fenestrate. Carotid single or double. Nature completely altricial, but young sometimes hatched with down[11] (e. g. Caprimulgidæ).
a. Cypseli. Palate ægithognathous. Wings lengthened in their terminal portions, abbreviated basally, with the first primary not reduced. Tail of ten rectrices. Bill fissirostral or tenuirostral. Feet never zygodactyle nor syndactyle, small, weak, scarcely fitted for locomotion; hallux often elevated or lateral or reversed; front toes usually webbed at base, or with abnormal ratio of phalanges in length and number, or both. Sternum deep-keeled, usually entire or else doubly notched or perforate. Syringeal muscles not more than one pair.
b. Cuculi. Palate desmognathous. Wings not peculiar in brevity of proximal or length of distal portions, and with first primary not reduced. Tail of eight to twelve rectrices. Bill of indeterminate form, never cered; tongue not extensile. Feet variously modified by versatility or reversion of either first, second, or fourth toes, or by cohesion for a great distance of third and fourth, or by absence or rudimentary condition of first or second; often highly scansorial, rarely ambulatorial. Syringeal muscles two pairs at most.
c. Pici. Palate “exhibiting a simplification and degradation of the ægithognathous structure” (Huxley); wings bearing out this passerine affinity in the common reduction of the first primary and the restriction of the greater coverts. Tail of ten perfect rectrices and usually a supplementary pair. Rostrum hard, straight, narrow, subequal to head, with commonly extensile and vermiform but not furcate tongue. Feet highly scansorial. Fourth toe permanently reversed; basal phalanges of toes abbreviated. Sternum doubly notched. Salivary glands highly developed. Hyoidean apparatus peculiar.