322 
are each tuned to a distinct frequency, 
like the strings of a harp, and that the 
basilar membrane is a_ frequency 
analyzer. The number of fibers in the 
pigeon is, however, only about one- 
eighth of the number in the human 
cochlea. 
At the apical end of the cochlea, 
beyond the end of the basilar mem- 
brane, there is a further range of hair 
cells whose hairs are embedded in a 
mucoid cap or cupola. The cap, unlike 
the tectorial membrane, is loaded with 
calcareous particles. Of this sense 
organ, the lagena, no trace remains in 
the human cochlea (or, indeed, in any 
mammal except the egg-laying Mono- 
tremes), but it is known that it is 
important in the hearing of fish and of 
many reptiles which have no basilar 
membrane. It is known, moreover, 
that these animals are less responsive 
to high-pitched sounds than are the 
mammals. Consequently it is a legiti- 
mate inference that birds have a dual 
auditory mechanism, the lagena being 
responsive to low frequencies and the 
cochlea proper, with its basilar mem- 
brane, to high frequencies. It may, 
indeed, be that the ranges of frequen- 
cies to which lagena and cochlea re- 
spond do not overlap. Small birds 
may have no interest in the band of 
frequencies lying between the thud of 
footsteps and their own shrill twitter- 
ing. But this is necessarily pure specu- 
lation, for no reliable physiological 
work has been done on the hearing of 
birds. ‘Though there are well-authen- 
ticated records of pheasants being dis- 
turbed by distant gunfire and explo- 
sions quite inaudible to man, it is not 
even certain whether it was the air- 
borne waves or the earth tremor which 
was the disturbing factor; and the 
upper frequency limit is also quite 
uncertain for any species of bird. 
What little there is to say is an infer- 
ence from behavior. It seems certain 
that birds can hear the voices of their 
own species, and at least probable that 
they can hear those of other species. 
Sc the auditory range may extend 
from the bark of a raven or the cooing 
ANNUAL REPORT SMITHSONIAN INSTITUTION, 1948 
of a pigeon to the cheep of a chiffchaff 
or, say, from 200 cycles/sec. to 10 
kilocycles. It is most unlikely, how- 
ever, that 10 kilocycles represents the 
upper limit for small birds since recog- 
nition of an individual song or call 
seems to require perception of at 
least the first few harmonics as well as 
the fundamental of the highest tone 
in the song. We may plausibly guess 
that the auditory spectrum of chiff- 
chaff and willow wren extends higher 
than that of man, perhaps to 30 or 
40 kilocyeles. 
The rather widespread faculty of 
mimicry in birds is another indication 
that they are capable of pitch and 
intensity discrimination comparable 
with human ability in these respects, 
and noticeable departures from verisi- 
militude are at least as likely to be due 
to limitations imposed by the syrinx 
and buccal cavities of the mimic as to 
its failure to appreciate what it hears. 
But as a matter of fact the performance 
of some exotic mimics is staggering. 
Australian lyrebirds and bowerbirds 
are said to imitate with deceptive 
accuracy the patter of dry leaves in a 
breeze, the twang of a man climbing 
a tightly strung wire fence, and the roll 
of distant thunder—all of which are 
sounds of a type which is reproduced 
with notorious infidelity by grama- 
phone and radio, Clearly, therefore, 
these birds have an auditory spectrum 
substantially wider than the pass band 
of a broadcast receiver. 
It is perhaps significant that in the 
collection of labryinths of birds of the 
late A. A. Gray the three mentioned as 
having aconspicuously elongated coch- 
lea were a cockatoo, a thrush, and an 
owl. The first is a noted mimic, the 
second an occasional mimic whose 
habit of repeating a theme without 
perceptible deviation points to accu- 
rate pitch discrimination, the third 
has other reasons (discussed below) for 
requiring a frequency analyzer of high 
performance. 
The labyrinths figured by Retzius 
can be grouped by inspection as fol- 
lows: 
