ORNITHOLOGY 45 of Families or genera usually referred to the former plainly have some features in common, the few Families or genera that have been clubbed together in the latter make an assemblage that is quite artificial, though it may be freely owned that with our present knowledge it is impossible to determine the natural alliances of all of them. 1 That our knowledge is also too imperfect to enable systematists to compose a phylogeny of Birds, even of the Carinate Subclass, and draw out their pedigree, ought to be sufficiently evident. The uncertainty which still pre vails among the best-informed ornithologists as to the respective origin of the Ratitx and Carinatse is in itself a proof of that fact, and in regard to some groups much less widely differentiated the same thing occurs. We can point to some forms which seem to be collaterally ancestral (if such a phrase may be allowed), and among them perhaps some of those which have been referred to the group " Alectorides" just mentioned, and from a considera tion of their Geographical Distribution and especially Isolation it will be obvious that they are the remnants of a very ancient and more generalized stock which in various parts of the world have become more or less specialized. The very case of the New-Caledonian Kagu (Rhinochetus), combining features which occasionally recall the Sun- Bittern (Eurypyga), and again present an unmistakable likeness to the Limicolse or the Rallidge, shews that it is without any very near relation on the earth, and, if con venience permitted, would almost justify us in placing it in a group apart from any other, though possessing some characteristics in common with several. It is anything but the desire of the present writer to invent a new arrangement of Birds. Such acquaintance as he possesses with the plans which have been already propounded warns him that until a great deal more labour has been expended, and its results made clearly known, no general scheme of Classification will deserve to be regarded as final. Nevertheless in the best of modern systems there are some points which, as already hinted, seem to be well established, while in them there are also some dispositions and assignments which he is as yet unable to accept, while he knows that he is not alone in his mistrust of them, and lie thinks it his duty here to mention them in the hope that thereby attention may be further directed to them, and his doubts either dispelled or established it matters not which. The most convenient way of bringing them to the notice of the reader will per haps be by considering in succession the different groups set forth by the latest systematist of any authority Mr Sclater a sketch of whose method has been above given. If we trust to the results at which Prof. Huxley arrived, there can be little doubt as to the propriety of beginning the Carinate Subclass with his Dromse-ognathae, the Crypturi of Illiger and others, or Tinamous, for their resemblance to the Ratitcv is not to be disputed ; but it must be borne in mind that nothing whatever is known of their mode of development, and that this may, when made out, seriously modify their position relatively to another group, the normal Anseres, in which the investigations of Cuvier and L Herminier have already shewn that there is some resemblance to the Rat it x as regards the ossification of the sternum. It will be for embryologists to determine whether this asserted resemblance has any real meaning ; but of the sufficient standing of the Cryptvri as an Order there can hardly be a question. 1 Heterogeneous as is the group as left by the latest systematist, it is nothing to its state when first founded by Illiger in 1811 ; for it then contained in addition the genera Glareola and Cereopsis, but the last was restored to its true place among the Anseres by Temminck. The Alectn cleft of Dumeril have nothing in common with the A lectorides of Illiger, and the. latter is a name most unfortunately chosen, since the group so called does not include any Cock-like Bird. and a We have seen that Prof. Huxley would derive all other existing Carinate Birds from the Dromseognatkse ; but of course it must be understood in this, as in every other similar case, that it is not thereby implied that the modern representatives of the Dromseognathons type (namely, the Tinamous) stand in the line of ancestry. Under the name Impennes we have a group of Birds, the Impennes. Penguins, smaller even than the last, and one over which until lately systematists have been sadly at fault ; for, though we as yet know little if anything definite as to their embryology, no one, free from bias, can examine any member of the group, either externally or internally, without perceiving how completely different it is from an} others of the Carinate division. There is perhaps scarcely a feather or a bone which is not diagnostic, and nearly every character hitherto observed points to a low morpho logical rank. It may even be that the clothing of Ilesper- ornis was not very dissimilar to the " plumage " which now covers the Impennes, and the title of an Order can hardly be refused to them. The group known as Pygopodes has been often asserted to be closely akin to the Impennes, and we have seen that Brandt combined the two under the name of Urinatores, while Mr Sclater thinks the Pygopodea " seem to form a natural transition between " the Gulls and the Penguins. The affinity of the Alcidse or Auks (and through them the Divers or Colymbidie) to the Gulls may be a matter beyond doubt, and there appears to be ground for considering them to be the degraded offspring of the former ; but to the present writer it appears questionable whether the Grebes, Podicipedidx, have any real affinity to the two Families with which they are usually associated, and this is a point deserving of more attention on the part of morphologists than it has hitherto received. Under the name of Gavise, the Gulls and their close allies form a very natural section, but it probably hardly merits the rank of an Order more than the Pyyopodes, for its relations to the large and somewhat multiform though very natural group Limicolse have to be taken into consideration. Prof. PARKER long ago observed (Trans. Zool. Society, . p. 150) that characters exhibited by Gulls when young, but lost by them when adult, are found in certain Plovers at all ages, and hence it would appear that the Gavise are but more advanced Limicolx. The Limicoline genera Dramas and Ckionis have many points of resemblance to the Laridse ; and on the whole the proper inference would seem to be that the Limicolse, or something very like them, form the parent-stock whence have descended the Gavige, from which or from their ancestral forms the Alcidse have proceeded as a degenerate branch. If this hypothesis be correct, the association of these three groups would constitute an Order, of which the highest Family would perhaps be Otididse, the Bustards ; but until further research shews whether the view can be maintained it is not worth while to encumber nomenclature by inventing a new name for the combination. On the other hand the Petrels, which form the group Tubinares, would seem for Tubinares. several reasons to be perfectly distinct from the Gavige, and their allies, and possibly will have to rank as an Order. Considerable doubt has already been expressed as to the " A lee to- existence of an Order Aledorides, which no one can regard as a natural group, and it has just been proposed to rctransfer to the Limicolte one of the Families, Otididse, kept in it by Mr Sclater. Another Family included in it by its founder is Cariamidee, the true |>lace of which has long been a puzzle to systematizers. The present writer is inclined to think that those who have urged its affinity to the Acripitres, and among them taxonomers starting from bases so opposite as Sundevall and Prof. Parker, have more nearly hit the mark, and accordingly would
