shewn (LARK, page 511), have the planta more divided than any other among the Oscines. It seems hardly possible to adduce anything that would more conclusively demonstrate the independent nature of each of these characters—the complicated structure of the syrinx and the asserted inferior formation of the planta—which are in the Alaudidæ associated.1 Moreover, this same Family affords a very valid protest against the extreme value attached to the presence or absence of the outermost quillfeather of the wings, and in this work it is also shewn (loc. cit.) that almost every stage of magnitude in this feather is exhibited by the Larks from its almost abortive condition in Alauda to its very considerable development in Mirafra. Indeed there are many genera of Oscines in which the proportion that the outermost "primary" bears to the rest is at best but a specific character, and certain exceptions are allowed by Prof. Cabanis (p. 313) to exist.2 Some of them it is now easy to explain, inasmuch as in a few cases the apparently aberrant genera have elsewhere found a more natural position, a contingency to which he himself was fully awake. But as a rule the allocation and ranking of the different Families of Oscines by this author must be deemed arbitrary. Yet the value of his Ornithologische Notizen is great, not only as evidence of his extensive acquaintance with different forms, which is proclaimed in every page, but in leading to a far fuller appreciation of characters that certainly should on no account be neglected, though too much importance may easily be, and already has been, assigned to them.4

This will perhaps be the most convenient place to mention another kind of classification of Birds, which, based on a principle wholly different from those that have just been explained, requires a few words, though it has not been productive, nor is it likely, from all that appears, to be productive of any great effect. So long ago as 1831, Bonaparte, in his Saggio di una distribuzione metodica degli Animali Vertebrati, published at Rome, and in 1837 communicated to the Linnean Society of London, 'A new Systematic Arrangement of Vertebrated Animals,' which was subsequently printed in that Society's Transactions (xviii. pp. 247-304), though before it appeared there was issued at Bologna, under the title of Synopsis Vertebratorum Systematis, a Latin translation of it. Herein he them below the Family called by him Sylvicolidæ, consisting chiefly of the American forms now known as Mniotiltidæ, none of which as songsters approach those of the Old World.

1 It must be observed that Prof. Cabanis does not place the Alaudida lowest of the seventeen Families of which he makes the Oscines to be composed. They stand eleventh in order, while the Corvidæ are last-a matter on which something may be said in the sequel.

2 The American Family Vireonidæ (VIREO) presents some notable examples, though there it is stated that the tenth primary is always present, but often concealed by the ninth (cf. Coues, Key N. Am. Birds, ed. 2, p. 331).

3 By a curious error, probably of the press, the number of primaries assigned to the Paradiseidæ and Corvidæ is wrong (pp. 334, 335). In each case 10 should be substituted for 19 and 14.

* A more extensive and detailed application of his method was begun by Prof. Cabanis in the Museum Heineanum, a useful catalogue of specimens in the collection of the late Oberamtmann Heine, of which the first part appeared at Halberstadt in 1850, and the last, the work being still unfinished, in 1863. A Nomenclator of the same collection was printed at Berlin 1882-90 by its owner's son and Dr. Reichenow.

divided the Class Aves into two Subclasses, to which he applied the names of Insessores and Grallatores (hitherto used by their inventors Vigors and Illiger in a different sense), in the latter work relying chiefly for this division on characters which had not before been used by any systematist, namely, that in the former group Monogamy generally prevailed and the helpless nestlings were fed by their parents, while the latter group were mostly Polygamous, and the chicks at birth were active and capable of feeding themselves. This method, which in process of time was dignified by the title of a Physiological Arrangement, was insisted upon with more or less pertinacity by the author throughout a long series of publications, some of them separate books, some of them contributed to the memoirs issued by many scientific bodies of various European countries, ceasing only at his death, which in July 1857 found him occupied upon the unfinished Conspectus Generum Avium before mentioned. In the course of this series, however, he saw fit to alter the name of his two Subclasses, since those which he at first adopted were open to a variety of meanings, and in a communication to the French Academy of Sciences in 1853 (Comptes Rendus, xxxvii. pp. 641-647) the denomination Insessores was changed to Altrices, and Grallatores to Præcoces-the terms now preferred by him being taken from Sundevall's treatise of 1835 already mentioned. The views of Bonaparte were, it appears, also shared by an ornithological amateur of some distinction, Hogg, who propounded a scheme which, as he subsequently stated (Zool. 1850, p. 2797), was founded strictly in accordance with them; but it would seem that, allowing his convictions to be warped by other considerations, he abandoned the original "physiological" basis of his system, so that this, when published in 1846 (Edinb. N. Philos. Journ. xli. pp. 50-71) was found to be established on a single character of the feet only, whereon he defined his Subclasses Constrictipedes and Inconstrictipedes. The numerous errors made in his assertion hardly need pointing out. Yet the idea of a "physiological" arrangement on the same kind of principle found another follower, or, as he thought, inventor, in Newman, who published (Proc. Zool. Soc. 1850, pp. 46-48, and Zool. pp. 2780-2782) a plan based on exactly the same considerations, dividing Birds into two groups, "Hesthogenous' -a word so vicious in formation as to be incapable of amendment, but intended to signify those that were hatched with a clothing of down—and Gymnogenous," or those that were hatched naked. These three systems are essentially identical; but, plausible as they may be at the first aspect, they have been found to be practically useless, though such of their characters as their upholders have advanced with truth deserve attention, and, as will be seen in the present work, Dr. Gadow's terms Nidicola and Nidifugæ, used in no systematic sense, express with greater accuracy what is needed. Physiology may one day very likely assist the systematist; but it must be real physiology and not a sham.

In 1856 Prof. Gervais, who had already contributed to the Zoologie of M. de Castelnau's Expédition dans les parties centrales de l'Amérique du Sud some important memoirs describing the anatomy of the HOACTZIN (page 421) and certain other Birds of doubtful or anomalous position, published some remarks on the characters which could be drawn from the

sternum of Birds (Ann. Sc. Nat. Zoologie, ser. 4, vi. pp. 5-15). The considerations are not very striking from a general point of view; but the author adds to the weight of evidence which some of his predecessors had brought to bear on certain matters, particularly in aiding to abolish the artificial groups "Déodactyls," "Syndactyls" and "Zygodactyls," on which so much reliance had been placed by many of his countrymen ; and it is with him a great merit that he was the first apparently to recognize publicly that characters drawn from the posterior part of the sternum, and particularly from the "échancrures," commonly called in English "notches" or "emarginations," are of comparatively little importance, since their number is apt to vary in forms that are most closely allied, and even in species that are usually associated in the same genus or unquestionably belong to the same Family,1 while these "notches," sometimes become simple foramina, as in certain Pigeons, or on the other hand foramina may exceptionally change to "notches," and not unfrequently disappear wholly. Among his chief systematic determinations we may mention that he refers the Tinamous to the Rails, because apparently of their deep "notches," but otherwise takes a view of that group more correct according to modern notions than did most of his contemporaries. The Bustards he would place with the "Limicoles," as also Dromas (CRAB-PLOVER) and Chionis, (SHEATHBILL). Phaethon (TROPICBIRD) he would place with the "Laridés" and not with the "Pelécanidés,” which it only resembles in its feet having all the toes connected by a web. Finally Divers, Auks and Penguins, according to him, form the last term in the series, and it seems fit to him that they should be regarded as forming a separate Order. It is a curious fact that even at a date so late as this, and by an investigator so well informed, doubt should still have existed whether Apteryx should be referred to the group containing the Cassowary and the Ostrich. On the whole the remarks of this esteemed author do not go much beyond such as might occur to any one who had made a study of a good series of specimens; but many of them are published for the first time, and the author is careful to insist on the necessity of not resting solely on sternal characters, but associating with them those drawn from other parts of the body.

Three years later in the same journal (xi. pp. 11-145, pls. 2-4) M. Blanchard published some Recherches sur les caractères ostéologiques des Oiseaux appliquées à la Classification naturelle de ces animaux, strongly urging the superiority of such characters over those drawn from the bill or feet, which, he remarks, though they may have sometimes given correct notions, have mostly led to mistakes, and, if observations of habits and food have sometimes afforded happy results, they have often been deceptive; so that, should more be wanted than to draw up a mere inventory of creation or trace the distinctive outline of each species, zoology without anatomy would remain a barren study. At the same time he states that authors who have occupied themselves with the sternum alone have often

1 Thus he cites the cases of Machetes pugnax and Scolopax rusticula among the "Limicoles," and Larus cataractes among the "Laridés," as differing from their nearest allies by the possession of only one "notch" on either side of the keel (cf. suprà, page 49).

produced uncertain results, especially when they have neglected its anterior for its posterior part; for in truth every bone of the skeleton ought to be studied in all its details. Yet this distinguished zoologist selects the sternum as furnishing the key to his primary groups or "Orders"" of the Class, adopting, as Merrem had done long before, the same two divisions Carinata and Ratitæ, naming, however, the former Tropidosternii and the latter Homalosternii.1 Some unkind fate has hitherto hindered him from making known to the world the rest of his researches in regard to the other bones of the skeleton till he reached the head, and in the memoir cited he treats of the sternum of only a portion of his first "Order." This is the more to be regretted by all ornithologists since he intended to conclude with what to them would have been a very great boon-the shewing in what way external characters coincided with those presented by Osteology. It was also within the scope of his plan to have continued on a more extended scale the researches on ossification begun by L'Herminier, and thus M. Blanchard's investigations, if completed, would obviously have taken extraordinarily high rank among the highest contributions to ornithology. As it is, the 32 pages we have of them are of considerable importance; for, in this unfortunately unfinished meinoir, he describes in some detail the several differences which the sternum in a great many different groups of his Tropidosternii presents, and to some extent makes a methodical disposition of them accordingly. Thus he separates the Birds-of-Prey into three great groups—(1) the ordinary Diurnal forms, including the Falconida and Vulturide of the systematist of his time, but distinguishing the American VULTURES from those of the Old World; (2) Gypogeranus (SECRETARY-BIRD); and (3) the OWLS. Next he places the Parrots, and then the vast assemblage of "Passereaux "-which he declares to be all of one type, even genera like Pipra (MANAKIN) and PITTA-and concludes with the somewhat heterogeneous conglomeration of forms, beginning with Cypselus (SWIFT), that so many systematists have been accustomed to call Picaria, though to them as a group he assigns no name.2

Important as are the characters afforded by the sternum, that bone even with the whole sternal apparatus should obviously not be considered alone. To aid ornithologists in their studies in this respect, Eyton, who for many years had been forming a collection of Bird's skeletons, began the publication of a series of plates representing them. The first part of this work, Osteologia Avium, appeared early in 1859, and a volume was completed in 1867. A supplement was issued in 1869, and a Second Supplement, in three parts, between 1873 and 1875. The whole work contains a great number of figures of Birds' skeletons and detached bones; but they are not so drawn as to be of much practical use, and the

1 These terms were explained in his great work L'Organisation du Règne Animal, Oiseaux (p. 16), begun in 1855, and unhappily unfinished, to mean exactly the same as those applied by Merrem to his two primary divisions.

2 M. Blanchard's animadversions on the employment of external characters, and on trusting to observations on the habits of Birds, called forth a rejoinder from Mr. Wallace (Ibis, 1864, pp. 36-41), who successfully shewed that they are not altogether to be despised.

A somewhat

accompanying letterpress is too brief to be satisfactory. similar work, Abbildungen von Vogel-Skeletten, was begun in 1879 by Dr. A. B. Meyer, and is still in progress, 210 plates of Birds' skeletons having already appeared. Some of these are excellent, but photography, by means of which they are all represented, is an unintelligent art, and as the sun shines alike on the evil and the good, so minor characters are as faithfully portrayed as those which are of importance, and indeed the latter are often, from the nature of the case, obscure or even indistinguishable. Yet we may be sure that every possible care was taken to avoid the disappointment thus caused.1

That the eggs laid by Birds should offer to some extent characters of utility to systematists is only to be expected, when it is considered that those from the same nest generally bear an extraordinary family-likeness to one another, and also that in certain groups the essential peculiarities of the egg-shell are constantly and distinctively characteristic. Thus no one who has ever examined the egg of a Duck or of a Tinamou would ever be in danger of not referring another Tinamou's egg or another Duck's that he might see to its proper Family, and so on with many others. Yet, as is stated in the text (p. 182), the expectation held out to oologists, and by them, of the benefits to be conferred upon Systematic Ornithology from the study of Birds' eggs, so far from being fulfilled, has not unfrequently led to disappointment. But at the same time many of the shortcomings of Oology in this respect must be set down to the defective information and observation of its votaries, among whom some have been very lax, not to say incautious, in not ascertaining on due evidence the parentage of their specimens, and the author next to be named is open to this charge. After several minor notices that appeared in journals at various times, Des Murs in 1860 brought out at Paris his ambitious Traité général d'Oologie Ornithologique au point de vue de la Classification, elsewhere mentioned (EGGS, page 191, note), which contains (pp. 529-538) a 'Systema Oologicum' as the final result of his labours. In this scheme Birds are arranged according to what the author considered to be their natural method and sequence; but the result exhibits some unions as ill-assorted as can well be met with in the whole range of tentative arrangements of the Class, together with some very unjustifiable divorces. This being the case, it would seem useless to take up further space by analysing the several proposed modifications of Cuvier's arrangement which the author takes as his basis. The great merit of the work is that the author shews the necessity of taking Oology into account when investigating the classification of Birds, but it also proves that in so doing the paramount consideration lies in the thorough sifting of evidence as to the parentage of the eggs which are to serve as the building stones of the fabric to be erected (Ibis, 1860, pp. 331-335). The attempt of Des Murs was praiseworthy; but in effect it has utterly failed, notwithstand

1 A countless number of osteological papers have appeared in journals, and to name them would here be impossible. The more important have generally been mentioned in the body of this work in connexion with the species or group of species they illustrate; but many that are good are necessarily passed over.