ORNITHOLOGY 47 possibly from some ancestral type akin to and now most directly represented by the enigmatical Cariama possibly in some other way which we can only dimly foreshadow. The Herodiones are commonly partitioned into three groups Ardex, Ciconise, and Platalex, the last including the Ibises which may certainly be considered to be as many Suborders. The second of them, the Storks, may perhaps be regarded as the point of departure for the Accipitres in the manner indicated, 1 as well as, according to Prof. Huxley, for the Flamingoes, of which he would make a distinct group, Ampkimorphoe, equivalent to the Odontoglossx of Nitzsch, intermediate between the Pelarffomorphte and the Chenomorphx, that is, between the Storks and the Geese. When the embryology of the P/icenicopteridx is investi gated their supposed relationship may perhaps be made oat. At present it is, like so much that needs to be here advanced, very hypothetical; but there is so much in the osteology of the Flamingoes, besides other things, that resembles the Aiiseres that it would seem better to regard them as forming a Subclass of that group to rank equally with the true Anseres and with the Pulamedeas (SCREAMER, </.? .), which last, notwithstanding the opinion of Garrod, can hardly from their osteological similarity to the true Anseres be removed from their neighbourhood. Whatever be the alliances of the genealogy of the Accipitres, the Diurnal Birds-of-Prey, their main body must stand alone, hardly divisible into more than two principal groups (1) containing the Cathartidx or the Vultures of the New World, and (2) all the rest, though no doubt the latter may be easily subdivided into at least two Families, Vulturidae. and Falconidse, and the last into many smaller sections, as has commonly been done ; but then we have the outliers left. The African Serpentariidx, though represented only by a single species, 2 are fully allowed to form a type equivalent to the true Accipitres composing the main body ; but whether to the Secretary-bird should be added the often-named Cariama, with its two species, must still remain an open question. It has so long been the custom to place the Owls next to the Diurnal Birds-of-Prey that any attempt to remove them from that position cannot fail to incur criticism. Yet when we disregard their carnivorous habits, and certain modifications which may possibly be thereby induced, we find almost nothing of value to indicate relationship between them. That the Striges stand quite independently of the Accipitres as above limited can hardly be doubted, and, while the Psittaci or Parrots would on some grounds appear to be the nearest allies of the Accipitres, the nearest relations of the Owls must be looked for in the multifarious group Picarise. Here we have the singular Steatornis (GrjACHARO, vol. ix. p. 227), which, long confounded with the Caprimulgidae (GOATSUCKER, vol. ix. p. 711), has at last been recognized as an indepen dent form, and one cannot but think that it has branched off from a common ancestor with the Owls. The Goat suckers may have done the like, 3 for there is really not much to ally them to the Swifts and Humming-birds, the Macrochires proper, as has often been recommended. However, the present writer would not have it supposed that he would place the Striges under the Picarise, for the 1 Garrod and Forbes suggest a "Ciconiiform" origin for the Tubinares (Zool. Voy. "Challenger," pt. xi. pp. 62, 63). ^It was long suspected that the genus Polyboroides of South Africa and Madagascar, from its general resemblance in plumage and outward form, might come into this group, but that idea has now been fully dispelled by M. A. Milne-Edwards in his and M. Grandidier s magnificent Oiseaux de Madagascar (vol. i. pp. 50-66). 3 The great resemblance in coloration between Goatsuckers and Owls is of course obvious, so obvious indeed as to make one suspicious of their being akin ; but in reality the existence of the likeness is no bar to the affinity of the groups ; it merely has to be wholly disregarded. last are already a sufficiently heterogeneous assemblage, and one with which he would not meddle. Whether the Woodpeckers should be separated from the rest is a matter of deeper consideration after the deliberate opinion of Prof. Parker, who would lift them as Saurognathee, to a higher rank than that in which Prof. Huxley left them as Celeomorphse, indeed to be the peers of Sckizognathx, Desmognatkse, and so forth ; but this advancement is based solely on the characters of their palatal structure, and is unsupported by any others. That the Pici constitute a very natural and easily defined group is indisputable; more than that, they are perhaps the most differentiated group of all those that are retained in the " Order " Picarise ; but it does not seem advisable at present to deliver them from that chaos when so many other groups have to be left in it. Lastly we arrive at the Passeres, and here, as already Passeres, mentioned, the researches of Garrod and Forbes prove to be of immense service. It is of course not to be supposed that they have exhausted the subject even as regards their J/esomyodi, while their Acromyodi were left almost untouched so far as concerns details of arrangement ; but the present writer has no wish to disturb by other than very slight modifications the scheme they put forth. He would agree with Mr Sclater in disregarding the distinc tions of Desmodactyli and Eleutherodaetyli, grouping the former (Euryleemidx) with the Heteromeri and Ilaploo- phonse, which all together then might be termed the their Sub- Suborder Oligomyodi. To this would follow as a second orders. Suborder the Tracheophonee as left by Garrod, and then as a third Suborder the abnormal Acromyodi, whether they are to be called Pseudoscines or not, that small group con taining, so far as is known at present, only the two Families Atrichiidse and Menuridse. Finally we have the normal Acromyodi or true seines. This last and highest group of Birds is one which, as Oscines, before hinted, it is very hard to subdivide. Some two or their homo- three natural, because well-differentiated, Families are to K C BCOUS - be found in it such, for instance, as the Hirundinidse or n Swallows, which have no near relations ; the Alaudidce or Larks, that can be unfailingly distinguished at a glance by their scutellated planta, as has been before mentioned ; or the Meliphagidse with their curiously constructed tongue. But the great mass, comprehending incomparably the greatest number of genera and species of Birds, defies any sure means of separation. Here and there, of course, a good many individual genera may be picked out capable of the most accurate definition ; but genera like these are in the minority, and most of the remainder present several apparent alliances, from which we are at a loss to choose that which is nearest. Four of the six groups of Mr Sclater s " Laminiplantar " Oscines seem to pass almost imperceptibly into one another. We may take examples in which what we may call the Thrush-form, the Tree- creeper-form, the Finch-form, or the Crow-form is pushed to the most extreme point of differentiation, but we shall find that between the outposts thus established there exists a regular chain of intermediate stations so intimately con nected that no precise lines of demarcation can be drawn cutting off one from the other. Still one thing is possible. Hard though it be to find Supposed definitions for the several groups of Oscines, whether we high rank of make them more or fewer, it is by no means so hard, if we go the right way to work, to determine which of them is the highest, and, possibly, which of them is the lowest. It has already been shewn (page 30) how, by a woe ful want of the logical apprehension of facts, the Turdidse came to be accounted the highest, and the position ac corded to them has been generally acquiesced in by those who have followed in the footsteps of Keyserling and
