as one unit, but just parts of them vibrate, and maybe that process is
helped by combining a thin vibrating process with the big inertial weight
of the rest of the ossicle. That would especially be useful for high-fre-
quency hearing—but this is all speculation. It is very difficult to study
movements of the ossicles in a cetacean.
It would appear that many of the specializations of hearing in mod-
ern whales are actually for high-frequency echolocation. Odontocetes
such as dolphins emit high-frequency sounds through specialized organs
140 | Chapter 11
3. The sounds are reflected off objects
around the whale, and the reflections
travel back to the whale.
2. A fat pad called the melon is located in the whale’s forehead.
It functions as an acoustic lens, focusing sounds as they travel
through the forehead.
blowhole
1. The whale produces sounds by
squirting air back and forth between
specialized organs that are part
of its nasal passages.
5. The fat pad ends at the
tympanic plate of the ear,
where sounds are passed
on to the inner ear
(see figure 41).
4. Reflected sounds are received
mandibular
by another fat pad inside the
foramen
lower jaw and passed to the ear. On their way
they traverse the mandibular foramen.
figure 42. The process of echolocation. The toothed whale (grey on the right) emits
sound waves from its forehead. These reflect off the fish, and the reflections are received
by the lower jaw and ear of the whale.
in their bulbous forehead and listen to the reflections of those sounds
from potential prey with their sophisticated ears (figure 42). As a result,
a blind dolphin can feed with little problem; a deaf dolphin will starve.
In modern odontocetes, the stiffness of the tympanic plate and the heavy
ossicles are adaptations for the perception of high frequencies, and not
simply adaptations for underwater hearing.
Confusingly, the ear anatomy of baleen whales, mysticetes, is similar
in many ways to that of toothed whales: the tympanic plate and heavy
ossicles, and the shape of the tympanic membrane. But mysticetes are
specialists at hearing low frequencies, not high ones. It is possible that
the ancestors of mysticetes were high-frequency hearers, and that they
retained some of the features of their ancestors but shifted others, to
tune the ear to low frequencies (figure 43).
The ear is a wonderful organ to study for a paleontologist, because
many of the important structures are bone and thus fossilize. For ceta-
ceans, changes in the mandibular foramen, tympanic plate, and ossicles
can all be studied in detail.5 The closest Eocene ancestors of mysticetes
and odontocetes are basilosaurids. They had a tympanic plate, a large
mandibular foramen, and heavy ossicles of the shape of modern whales,
and their tympanic membrane had the umbrella shape of their modern
relatives. It is also clear that they were not echolocators, since they do not
have the forehead organs needed to make echolocating sounds. It is likely
y
eatus
portant
?
ysticeti
g
e
skull
reception
-frequenc
ry mechanism
M baleen whales
Low
hearin
eatus lost
ater-borne
y m
ly developed, unim
W bone conduction
andibular fat
Echolocation
tional external auditory m
ost significant
eak
M pad-based system
dontoceti
M
Present as accesso
W
Absent or nearly so
O toothed whales
Mechanism of sound
panic plate
ternal auditor
Tym
enlarged
Loss of func
Tympanic ring reduced in siz Ex
ore isolation of ear region from
Basilosauridae
d
d
skull
Even m
Protocetidae
skull
s
Bars on top summarize the evolution of sound-
en further enlarge
ther isolation of
am
d
Fur
ear region from
-shaped tympanic membrane
c
andibular for
tial isolation of ear region from
tion
tion
Remingtonocetidae
M
Tympanic plate thinned and enlarge
Cone
Fully rotated ossicle
Ossicles enlarge
Par
ater-borne W
Substrate-borne bone conduc
bone conduc
contact betweeny
Ambulocetidae
Bon
mandible and tympani
andibular foramen enlargedM Mandibular wall thinned
yostotic
tly rotated
icetidae
Pak
Ossicles par
Ossicles pach Involucrum
Crura of incus similar in length
tyls
Rotational lever arm system
Cladogram showing evolution of features related to hearing.
terrestrial
tiodac
on land
r e t a
w n i
ar
e 43.ru
50 million years ago
40 million years ago
fig
transmission mechanisms.
142 | Chapter 11
5. Remingtonocetus, hearing in air
The incus, working with the eardrum in air,
is similar in size to that of seals, suggesting that
they may have similar hearing mechanism
4. Modern seals (phocids)
Malleus and incus are very heavy.
6. Remingtonocetus,
This may help in hearing
hearing underwater
1
underwater via
The mal eus, working with
gram
bone conduction
the tympanic plate underwater,
s
is similar in size to that of modern
whales, suggesting they have
0.1
3. Pakicetus
a similar hearing mechanism
gram Incus and eardrum
match those of
2. Modern whales and dolphins
modern land
Here too, the weight of the mal eus
10
mammals
and incus increases as the sound receiving
mil igram of
the same
area, the tympanic plate, increases
size
eight of malleus and incu
1
W
1. Land mammals
mil igram
The weight of mal eus and incus
increases as the eardrum increases in size
0.1
and this correlates to body size too
mil igram
10 mm 2
100 mm2
1000 mm2
10,000 mm2
Area of eardrum (or of tympanic plate for whales hearing underwater)
figure 44. Mass of the malleus (hammer) and incus (anvil) of
modern mammals and some fossil whales plotted against size of the
sound-input area of the skull (eardrum in air, tympanic plate
underwater in whales). Remingtonocetus may have had two
sound-transmission mechanisms, one for airborne and one for
waterborne sound. After Nummela et al. (2007).
that basilosaurids were specialized for high-frequency hearing, which is
consistent with the idea that mysticetes had high-frequency ancestors.
All of these insights, inconsistencies, and opportunities dance through
my head as I scrutinize the new pakicetid skulls. They have an involu-
crum like modern whales, but lack a large mandibular foramen and retain
the external auditory meatus, which is also present in land mammals. The
only ossicle we have, the incus, is heavier than that bone in land mam-
mals, but lighter than whales, and looks different from, well, every other
mammal incus (figure 44).
In air, pakicetids probably used the same sound-transmission mecha-
nism as land mammals do: sounds make the eardrum vibrate and cause
the ossicles to rattle. Underwater, it is likely that that system did not
work very well. Instead, pakicetids may have heard by means of a sound-
transmission mechanism called bone transmission, which does not allow
for directional hearing. Humans experience bone transmission, for
instance, when they are near loud, low-frequency sounds: the bass in a
rock concert will send many of its vibrations through the floors and
stands, and these reach the ear by passing through the person’s body, not
the air. Crocodiles lay their jaws on the ground and pick up the footsteps
The River Whales | 143
of their prey in that way,6 and mole rats push their jaws against the walls
of their tunnels to listen to sounds produced by animals in nearby tun-
nels.7 Some forms of bone conduction are aided by the presence of heavy
ossicles, and this may be the reason for the increased weight of pakicetid
ossicles. From there, it may have been passed on to pakicetid descend-
ants, including modern whales. Having said that, it is unlikely that paki-
cetids heard very well underwater, and they certainly could not distin-
guish where a bone-conducted sound came from.
Fossilized ears are also known for remingtonocetids. In this group
(and also the protocetid whales, which will be discussed in chapter 12),
the mandibular foramen is enlarged, the fat pad and tympanic plate are
present, and the ossicles are large, similar to modern whales. However,
these whales retain an external auditory meatus. These whales could
still hear in air, but the heavy ossicles must have made efficient transmis-
sion of faint sounds difficult.8 The mandibular fat pad was the sound
transmitter underwater, just as in modern odontocetes. This new sound-
receiving mechanism would make it possible for these Eocene whales to
hear directionally underwater, as long as the pathways of bone conduc-
tion were switched off and could not interfere with the mandibular
sound path. Bone conduction depends on a tight connection between
the organ of hearing and the rest of the body, and such a connection is
present in land mammals, as well as in pakicetids. But after pakicetids,
that connection changes. The connection of the bones of the ear is looser
in remingtonocetids than in pakicetids. In the former, a space occurs
between the bones that hold the middle ear and cochlea (the tympanic
and petrosal bones) and the rest of the skull. This space is larger in basi-
losaurids and later whales, and in modern dolphins and their relatives
the space is so large that the ear bones tend to fall out of the skull when
the soft tissues are removed. Moreover, in modern whales, that space is
an air-filled cavity, similar to the sinuses in a person’s forehead. That air
is an acoustic insulator: it does not let bone-conducted sound pass to
the ear. Undoubtedly, bone-conducted sound could cross to the ear in
remingtonocetids, but the beginnings of the acoustic isolation that pro-
vides directional underwater hearing in modern whales are there too.
Not much is known about the ears of Ambulocetus. There is only
one individual for the species for which the ears are preserved, and they
are damaged by fossilization. However, it is clear that the species did
have a partly enlarged mandibular foramen (figure 25) and a thin man-
dibular wall, both of which are involved in sound transmission through
the jaw.9 Most intriguing about Ambulocetus is that the jaw joint is
144 | Chapter 11
expanded in such a way that the mandibular condyle (the part of the
lower jaw that makes that joint) is in direct bony contact with the
tympanic bone. That direct connection could also be a path for sound
from jaw to ear, as it also occurs in mole rats. Ambulocetus may have
been an early experiment to involve the lower jaw in sound transmis-
sion—far from perfect, but better than what pakicetids had—but if so,
it was then quickly discarded in the evolutionary process with reming-
tonocetids.
Taken together, the ear story is intricate and exciting. Modern whales
have ears that are relatively similar, well adapted for underwater hear-
ing. The early whales show that hearing gradually changed and that
there was an experimental phase, where the sound-transmission mecha-
nism initially built for hearing in air was modified to allow bone-con-
ducted hearing, an imperfect system, before a new sound-transmission
mechanism evolved that was only perfected in early odontocetes. After
that, the original land-mammal system was lost.
pakicetid whales
The ears of pakicetids already suggest that they spent time in water; so
if, in Jurassic Park fashion, we could bring one back and put it in a zoo,
we had better keep that i
n mind (figure 45). On land, visitors would
think a pakicetid was a wolf with a long nose and an oddly long and
powerful tail (figure 46). Differently from wolves, though, we would
watch them in the underwater viewing area, since they would spend
much of their time wading in the water, spying over the water-line for
unsuspecting and thirsty prey.
These earliest of whales all lived in a geographically small area,10 in
what is now northern Pakistan and western India (figure 22), around
forty-nine million years ago. Just three genera are known: wolf-sized
Pakicetus and Nalacetus, and fox-sized Ichthyolestes. Himalayacetus
from India was also described as a pakicetid, but is more likely to be an
ambulocetid. Locality 62 in the Kala Chitta Hills has produced more
pakicetids than all other localities combined, but the site is a big jumble of
the bones of many individuals; there never has been an associated skele-
ton of a single individual, so the reconstructions are composites (figure
38). Ichthyolestes’s small size helps in distinguishing its bones from those
of the larger pakicetids. Pakicetus and Nalacetus teeth and tympanic
bones are different in shape, but their limb bones are difficult to distin-
guish.11
figure 45. Life reconstruction of Pakicetus, the first known whale. It is at the base of the cetacean radiation and lived forty-nine million years ago in what is now Pakistan.
Externally very different from modern whales, dolphins, and porpoises, it was an
amphibious wader that lived in shallow streams.
146 | Chapter 11
figure 46. The skeleton of the Eocene whale Pakicetus. The soccer
ball is 22 cm (8.5 inches) in diameter.
Feeding and Diet. A lot has been learned about pakicetid feeding in
recent years, but many questions remain. Stable-isotope studies show
that they drank freshwater and were flesh eaters,12 and they have sturdy
high-pointed front teeth, as is common in predators that grasp strug-
gling prey. The premolars are triangular, and upper and lower premolars
The Walking Whales Page 22