--The neotropical genus Harpagus is
usually placed with the kites. It is characterized by two "teeth," on the upper mandible. The scientific
names of the two species of the genus, diodon and bidentatus, both reflect this fact. The latter, which
is the more widespread of the two, is called the Double-toothed Kite. Such "teeth," which are doubt.
less adaptations for dismembering food, are not unusual in kites. Thus the species f the Old World
genus Aviceda have variously one or two such dentides, whereas the American Gray-headed Kite of
the genus Leptodon, a close ally of .4viceda, has one.
Some early authors, such as Sharpe in the "Catalogue of Birds in the British Museum," placed
these toothed kites in the Falconidae, but it was soon realized that they have nothing to do with that
family.
Hapagus is less obviously a kite than are Aiceda or Leptodon. In proportions and color it is
somewhat like many accipiters. Loye Miller (Condor, 39, 1937:219-221) in a paper on the affinities of
Harpgus remarked: "the general appearance of the bird when I picked it up was that of an Acclpiter
with an abbreviated tail .... "Professor Miller prepared a skeleton of this bird which he compared
with those of a White-tailed Kite (Elanus), a Cooper Hawk (Acdp/ter cooperS), and a Merlin (Falco
columbarius). He found Harpagus very different from Falco in most of the characters compared. On
the other hand, in several respects he found the osteolugy of Harpagus more like .4cciplter than like
Elanus. These data have recently been used to bring the systematic position of the genus again into
question. Hellmayr and Conover (Cat. Birds Amer., pt. 1, no. 4, 1949:32, footnote), although leaving
Harpgus in the "Milvinae," state: "According to Miller skeletal characters of this form reveal its
close kinship to Accipiter."
I am not competent to discuss Miller's osteological comparisons in detail but would merely sug-
gest that the similarities to Acclpiter and the differences from Elanus may be superficial adaptations.
In one comparison this is clearly stated to be the case: Elanus has a type of humerus usual in sailing
birds; the other two do not. I feel that if the comparison had been with a kite of more or less similar
proportions the results might have been very different. Gurney (Ibis, 1881:118-124) in a paper on
Harpagus states: "Mr. Ridgway informs me 'it is most nearly related, osteologically, to Ictila.'"
The latter, of course, like Elanus, is a long-winged gliding kite, but at least it belongs to the same
"subfamily" of kites as does Harpagus, assuming, as I believe to be the case, that the latter is a kite.
In any event, Miller was not at all emphatic about his conclusions. He merely listed certain similarities
between Harpagus and Acclpiter without implying that its status was finally settled. Gurney also noted
that Nitzsch had found the pterylugraphy of Harpagus to be similar to that of Baza (= A/ceda).
In a paper just published, Vesta and Erwin Stresemann (Jour. f. Ornith., 101, 1960:394) state that
they found the wing molt of Harpagus to be typical of the kites of the railvine group. They also note
that Suschkin had reached the same conclusion about the position of the genus in his work on the
diurnal raptores.
The "reputation" of Harpagus has been influenced by some field notes of the junior author of
"The Birds of the Santa Marta Region of Colombia" by W. E. Clyde Todd and M.A. Carriker, Jr.
(Ann. Carnegie Mus., 14, 1922:146), where it is portrayed as a very savage little hawk that "feeds
entirely upon small birds." We are not told the source of the evidence for this rather sweeping state-
ment beyond the fact that a captive specimen ate small, dead birds. Knowledge of this hawk was
enhanced by Laughlin's account (Condor, 39, 1937:137-139) of a pair that nested on Barro Colorado
Island, Panama. He refuted the bird-eating habits and found that Harpagus feeds upon large insects
and lizards, securing the latter by a chase "up a slanting branch, hopping after it--not flying--with
wings spread to maintain balance. This method of hunting, although it looked clumsy, was apparently
successful .... "I have found four other references to the food of Harpagus in the literature or on
specimen labels. All refer to insects; one refers also to reptiles.
Mr. Eugene Eisenmann, who has had considerable experience with Harpagus, including observa-
tions of the pair studied by Laughlin, tells me that small birds usually are unconcerned when a mem-
ber of this genus is nearby. Wetmore (Proc. U.S. Nat. Mus., 87, 1939:184) was made aware of the
presence' of Harpagus by the alarmed chirping of a large hummingbird, Cyanolesbia, and Eisenmann
saw one chased off by a pair of Streaked Flycatchers (Myiodynastes macu/atus). Hummingbirds and
tyrant flycatchers are, however, unusually pugnacious birds.
The problem of possible mimicry among birds of prey is one which still awaits clarification. Tbe
following paragraph from Newton's article on this subject in the "Dictionary of Birds" (p. 574) is
of interest: "On the information of Mr. Salvin, Mr. Wallace has cited (Contrib. Nat. Sdect. p. 107)
another very curious case of Mimicry in Birds. This is furnished by Acdpter peatus, a widely-ranging
species of Sparrow-Hawk which near Rio Janeiro departs from the plumage it wears in other
places to assume that of Harpagus diodon, a local species of insectivorous habit, with the object, as
suggested, of deluding small birds into the belief that it is harmless in character. The similarity here
extends to both immature and adult plumages, which are very different." It should be pointed out
that although the adult and immature plumages of Acdpiter bicolor pieatus are very like those of
Hartgus diodon, it is not, of course, definite that the former has departed from its usual coloration
only in the range of diodon or that the resemblance has any special significance. Races of bi,olor with
streaked immatures are found beyond the range of diodon; conversely, the northern half of the range
of diodon has another race of Accpiter in which the iramatures are immaculate white ventrally and
quite unlike those of diodon.
A further point is that, assuming there is mimicry involved, we do not know whether the more
predatory species (Acclpter) is mimicking the inoffensive one (Harpagus) or vice versa. As Meyer
and Wiglesworth (Birds of Celebes, 1, 1898:66-72, color pls. 2 and 3) state in their lengthy discus-
sion of the remarkable similarity in all plumages between a honey-buzzard kite (Pernis celebensis)
and the eagle (Spzaetus lanceolatus) it is possible that the inoffensive kite might obtain relief from
enemies, presumably nest predators, by resembling the eagle. They conclude, however, that both
species are so uncommon that it is difficult to postulate mimicry upon any basis whatever.
If Harpagus does mimic accipiters, this was of little avail in the pair observed by Laughlin. One
of the pair was rather easily frightened from the nest by a toucan (Rampkastos) which proceeded
to eat the kite's egg!
In summary, one may conclude that Hartgus is not related to Accipter and that the resemblance
between the two may possibly, although this is unlikely, represent mimicry.--DEAN AAOO, Ameri-
can Museum of Natural History, October 31, 1960.