that four may be accepted as its proper quota. The mistake then
chargeable to the above authors is that in their discussion they
transferred one toe from before and added it behind. In this curious way
came their zygodactylous bird.

This same pipe is figured by Stevens in Flint Chips, p. 426, Fig. 5. The
wood-cut is a poor one, and exhibits certain important changes, which,
on the assumption that the pipe is at all well illustrated by the cast
in the Smithsonian, reflects more credit on the artist's knowledge of
what a toucan ought to look like than on his fidelity as an exact
copyist.

The etchings across the upper surface of the base of the pipe, miscalled
fingers, are not only made to assume a hand-like appearance but the
accommodating fancy of the artist has provided a roundish object in the
palm, which the bird appears about to pick up. The bill, too, has been
altered, having become rounded and decidedly toucan-like, while the tail
has undergone abbreviation, also in the direction of likeness to the
toucan. In short, much that was lacking in the aboriginal artist's
conception towards the likeness of a toucan has in this figure been
supplied by his modern interpreter.

[Illustration: Fig. 17.--Toucan as figured by Stevens.]

This cut corresponds with the cast in the Smithsonian collection, in
having the normal number of toes, four--three in front and one behind.
This departure from the arrangement common to the toucan family, which
is zygodactylous, seems to have escaped Stevens's attention. At least he
volunteers no explanation of the discrepancy, being, doubtless,
influenced in his acceptance of the bird as a toucan by the statements