EXTERNAL PAHTS OF BIRDS. — THE FEET. 
127 
inner one. lu nearly every three-toed bird, all three toes are anterior ; our single exception is 
in the genus Picdides, where the true hind toe is wanting, the outer anterior one being reversed 
as usual in zygodactyles. No bird has more toes behind than in front. Birds' toes, and their 
respective joints, are 
Numbered, in a certain definite order, as follows (see figs. 34, 36) : hind toe = first toe, 
1^; inner anterior toe = second toe, 2t; middle anterior toe = third toe, St; outer anterior 
toe = fourth toe, 4 1. Such identification of 1 ^, 2 3 ^, 4 * applies to the ordinary case of three 
toes in front and one behind. But, obviously, it holds good for any other arrangement of the 
toes, if we only know M'hich one is changed in position, — a thing always easy to learn, as we 
shall see at (mce. In birds with the hind toe reversed, leaving all four in front, the same 
order is evident, though then 1 ^ is the inner anterior, 2 t the next, etc. ; for it always happens, 
when a hind toe turns forward, that it turns on the inner side of the foot. Similarly, in yoke- 
toed birds (excepting Trogonidce), it is the outer anterior which is turned backward, as above 
said ; then, evidently, inner hind toe = It ; inner front toe = 2 ^ ; outer front toe = 3 ^ ; outer 
hind toe = 4: t. In Trogonidce, with inner front toe reversed, the correction of the formula is 
easily made. Moreover, when the number of toes decreases from four to three or two, the 
digits are almost always reduced in the same order : thus, in three-toed birds, 1 ^ is the missing 
one ; in the two-toed ostrich, 1 1 and 2 1 are gone. The only known exceptions to this general- 
ization are afforded by two exotic genera of kingfishers, Ceyx and Alcyone, in which 2 ^ is 
defective ; and by the anomalous passerine Cholornis of China, in which 4 ^ is in like case. 
The rule is proven by the 
Number of Phalanges, or joints, of the digits. The constancy of the joints in birds^ 
toes is remarkable, — it is one of the strongest expressions of the highly monomorphic character 
of Aves. In all birds, excepting Procellariidce, 1 1 when present has two joints (not counting, 
of course, the accessory metatarsal). In all birds, 2 t when present has three joints. In nearly 
all birds, 3 1 has four joints. In nearly all birds, 4 t has five joints. Thus, any digit has one 
more joint than the number of itself. The exceptions to this regularity consist in the lessening 
of the number of joints of 1 ^ or 3 ^ by one, and of 4 f by one or two. So when the joints do 
not run 2, 3, 4, 5, for toes 1 to 4, they run either, 1, 3, 4, 5, or 2, 3, 4, 4, or 2, 3, 3, 3. (These 
statements do not regard the anomalous cases of Ceyx, Alcyone, and Cholornis — see above.) 
This variability is nearly confined to certain Picarian birds : our examples of it are in certain 
a genera of Cypselince, fig. 40, where the ratio is 2, 3, 3, 3, 
of CaprimulgincB, fig. 41, where it is 2, 3, 4, 4 ; and the petrel 
family, with 1, 3, 4, 5. Such admirable conservatism enables 
us to teR what toes are missing in any case, or what ones are 
^ out of the regular position. Thus, in Picdides, the hind toe, 
(r apparently 1 1, is known to be 4 t, because it is five-jointed ; 
Fig 40 Pha trogon, the inner hind toe is 2 t, being three -jointed ; in 
lange.s of Cypse- ostrich, with only two toes, 3 t and 4 1 are seen to be 
line foot, 2,3,3,3. preserved, because they are respectively four- and five-jointed. 
(See fig. 34, where the digits and their phalanges are numbered.) Besid^^ 
this interesting numerical ratio, the phalanges have other inter-relatio'^- of 
some consequence in classification, resulting from their comparative l<iigths. langesof Caprimul- 
In some families of birds, one or more of the lasal or proximal rialanges gine foot, 2, 3, 4, 4. 
(those next to the foot — opposed to distal, or those at the ends oUhe digits) of the front toes, 
are extremely short, being mere nodules of bone (fig. 40) ; in oher and more frequent cases^ 
they are the longest of all, as in figs. 34, 41. On the whole, t2ey generally decrease in length 
from proximal to distal extremity, and the last one of any 1f)e is quite small, serving merely 
