“Yeah! But where is the rest of it?”
We both frantically search for more in the bag of pieces.
Geological age-wise, our new whale falls between Ambulocetus and
the basilosaurids. The sacrum of those two is very different: four firmly
fused vertebrae in the former, no fused sacral vertebrae at all in the lat-
ter. That makes the difference between being able to support one’s
weight on land and not, so this is very important in understanding our
new whale. Ellen and I keep on going through our fossils, finding more
pieces. I glue them onto the growing sacrum with white Elmer’s glue. It
takes a while to dry, and I am impatient, looking and trying to fit more
fragments before already-glued joints have dried. Some come apart
from my handling the fossil. Ellen notices my impatience and takes the
sacrum from me without saying anything, firmly but gently. I know bet-
ter than to dissent. She is the fossil preparator. Things progress more
slowly, but now we need to glue each joint only once. Eventually, she fits
a nearly complete sacrum together: four fused vertebrae, with a large
joint for the pelvis. This guy could certainly stand on land.
Still high on the success of our reconstructive surgery on the sacrum,
I keep sorting through our bags of fossil chips. My eye is drawn toward
a bone the size of a toffee, and also that color. It is broken on three of
its four sides, indicating that it was part of something bigger, and the
breakage also explains why I ignored it before. However, the fourth side
shows two holes; they’re for teeth! A shiver runs over my back. “Sunil,
I’ve got a jaw.”
He looks up. The cavities that hold the left and right teeth, the alve-
oli, are very close together, indicating that the jaw is very narrow. We
both know what this means: this is one of the narrow-snouted whales
for which we have fragmentary skulls. It allows us to identify the speci-
men as a remingtonocetid.
Eventually, this new whale will receive the name Kutchicetus min-
imus, the “smallest whale from Kutch” (figure 31).1
Over time, too, it becomes clear what the big Y was: the impression
of the underside of the lower jaw. The long stem is the part where the
left and right jaw touch; the short arms are the left and right parts as
they diverge. The flute look-alike from Babia Hill, mentioned in chapter
6, would make just such an impression, and, indeed, they represent the
same species.
As Kutchicetus becomes well known, a company that makes museum
exhibits, Research Casting International, puts it all together and casts
it in some fancy plastic, for use in museum displays. Peter May, its
The Otter Whale | 107
figure 31. Life reconstruction of the Eocene whale Kutchicetus minimus, which lived
in India around forty-two million years ago. Kutchicetus and other remingtonocetids
were probably fish eaters, and were able to walk around on land.
director, and I get all the bones together, and his team makes mirror
images of the bones for which we only have one side. For the feet, there
is nothing to mirror image because we found no fossils of them. I do
not want the feet reconstructed, because I do not know what they
looked like. We decide that he will use wire to indicate where the toes
were, thus leaving it very clear that we do not have those parts.
108 | Chapter 8
Then I ask Carl Buell, a scientific illustrator, to make a drawing of the
animal. I know Carl. He is picky and precise. He asks for details—this
view, that bone, things that I have never thought about.
“How are the lips, are they floppy like a dog, or tight like a whale?”
Carl asks in his crackling voice that betrays that he is not a young man.
Carl knows his stuff—anatomy, function—he is also passionate. I say
that I do not know the answer to his question.
“It has this long, narrow snout, that’s crazy, I’d love to wrap my pen-
cils around that—a whale looking like a gharial.” He is right. This whale
does look like the narrow-snouted crocodilians that inhabit India and
Pakistan.
“So what do its feet look like?”
“We don’t have the feet. We didn’t find them.”
It is quiet for a second. He’s clearly disappointed.
“You don’t have any bones from the feet—nothing, no phalanges, no
“I have half a bone, probably a phalanx, it won’t help you.”
Quiet again. I break the silence. “What are you going to do with it in
the reconstruction?” I ask, somewhat worried that he may not think the
project is feasible.
“Oh, I’ll figure out something.” This does not sound good. I want to
know what is he planning.
“I don’t want you to make them up, OK?”
“You’ll see, you’ll like it.”
His voice changes to the tone of a doctor who comforts a worried
patient, but does not want to explain. I trust Carl, so I don’t push
him. He wants to know what to use as a scale, so people will know
how small it was. He proposes a shorebird. People have some sense of
what size those are, and Carl knows that they were around in the
Eocene.
A few days later, Carl sends me some sketches, the head from dorsal
and from the side—he is clearly struggling with it, it is so different from
other mammals—but no reconstruction of the body. I still do not know
how he is going to solve the hand-and-foot problem, and I am dying to
know. Then, eventually, I get a sketch of the entire animal. I rush to open
the file. There are three individuals, showing three different views of the
head, and he’s put them at the shoreline, feet in the water, invisible, the
rest of the animal above the water, visible. Brilliant, such a simple solu-
tion, except that I would not have thought of it. The bird in the back-
ground looks good, too.
The Otter Whale | 109
remingtonocetid whales
Carl’s reconstruction put some flesh on the bones of the remingtonocetid
whales, initially discovered by Ashok Sahni and his student V. P. Mishra.
After that initial description,2 Sahni sent another student, Kishor Kumar,
to Kutch to collect more whales. Bad weather made fieldwork impossi-
ble for much of Kishor’s stay, but he did find the most complete skull of
Remingtonocetus known at that time.3 He also collected new material
for another whale that Mishra had discovered: Andrewsiphius, named
after C. W. Andrews, a British paleontologist who worked in Egypt and
described many basilosaurids. Realizing that Andrewsiphius and Rem-
ingtonocetus were part of a unique Indo-Pakistani radiation of whales,
Kumar and Sahni combined Remingtonocetus and Andrewsiphius into
a new family: Remingtonocetidae. Since then, no remingtonocetid has
ever been found outside of the Indian continent, but three additional
genera have been described: Dalanistes, based on specimens from cen-
tral Pakistan4 and Kutch,5 Kutchicetus from Kutch, and Attockicetus,
which came from the same rocks as Ambulocetus: the Kuldana Forma-
tion of Northern Pakistan.6
Remingtonocetids differ from other Eocene cetaceans in having long
snouts, tiny eyes, and big ears. Based on what is known for Kutchicetus,
its body looked like that of an otter: short legs, and a long, powerful
tail. In contrast, the long, narrow snout makes the head look more like
a gharial (figure 33). Kutchicetus was the smallest remingtonocetid, the
size of a sea otter; Dalanistes was the largest, weighing maybe as much
as a male sea lion. Indian remingtonocetids are known from rocks forty-
two million years old.7 Attockicetus is older, the same age as Ambuloce-
tus, approximately forty-eight million years, and the remingtonocetids
from central Pakistan are between thirty-eight and forty-eight million
years old.8
Feeding and Diet. It is easy to imagine that the long snout helped rem-
ingtonocetids catch fish. If Ambulocetus lived like a crocodile, capturing
large, struggling prey, remingtonocetids were more delicate, lashing out
quickly with their sharp teeth when a fish came close.9 The front teeth of
Kutchicetus are long and slender, good for piercing and checking slippery
prey in a dash, but not for holding powerful struggling prey. The molars
are small, but tooth wear shows that Remingtonocetus chewed its food,
unlike modern whales; its teeth worked like those of basilosaurids and
ambulocetids. These molars cut like scissors, with sharp shearing edges
figure 32. Life reconstruction of the remingtonocetid whale Kutchicetus. Fossils of
hands and feet were not discovered for this whale, which means that the artist
reconstructing the animal needs to be creative.
The Otter Whale | 111
figure 33. The skeleton of the Eocene whale Kutchicetus minimus. Soccer ball is
22 cm (8.5 inches) in diameter.
(figure 34). No part of these teeth is involved in crushing food, unlike
Ambulocetus. Analysis of the stable isotopes of the teeth is consistent
with a fish diet, and further study may refine this.
The flute look-alike mentioned in chapter 6 is a Kutchicetus jaw from
Babia Hill. All its teeth fell out after death but before burial,10 but count-
ing sockets for the teeth (the alveoli) reveals what the dental formula is.
As in most early whales, there are three incisors, one canine, four premo-
lars, and three molars in both upper and lower jaw: 3.1.4.3/3.1.4.3.
Some jaws for Remingtonocetus and Dalanistes do still have teeth, and
surprisingly, the lower molars look like those of basilosaurid whales
(figure 34), with multiple cusps of decreasing size lined up from front to
back.11 However, these teeth are unlike basilosaurids’ in that they are
slender and delicate, not built to mince tough or hard food. In Andrews-
iphius, there are three low and flat cusps on a lower molar, lined up in a
row, with the middle cusp barely higher than the others. It is not clear
whether this unusual shape somehow related to a specialized function.
Remingtonocetid molars are specialized, unlike those of ambu-
locetids, which retain the shape of archaic land mammals with a high
front (the trigonid) and a low back (the talonid). The premolars in rem-
ingtonocetids have simple, triangular cusps. In Andrewsiphius and
Remingtonocetus, most premolars have two roots, but in Kutchicetus
there is only one root per tooth. In modern whales, there is never more
than a single root per tooth.
The most unusual feature of the jaws of remingtonocetids is the long
area of contact between left and right lower jaw. This area is called the
mandibular symphysis (figure 25). In ambulocetids and basilosaurids, the
left and right lower jaws are connected by ligaments; there is no bony
fusion across the mandibular symphysis. This is also the case in most
Remingtonocetus, although there is a bony connection in old individuals.
In Andrewsipius and Kutchicetus, the left and right jaws are joined by
Toward upper teeth
Cheekside of jaw (labial)
(occlusal)
LEFT LOWER MOLAR AS To front of jaw
To front of jaw
SEEN FROM OUTSIDE
(anterior)
(anterior)
(LABIAL VIEW)
LEFT UPPER MOLAR IN OCCLUSAL VIEW
Artiodactyla: Indohyus
Four large cusps, Four large cusps,
val eys between
one small cusp,
cusps.
val ey between
cusps.
RR 102
RR 209
Pakicetid whale
Only two large cusps One main cusp lost,
Pakicetus
and one small remain one small cusp lost,
valleys gone.
valley smaller.
H-GSP
Molar morphology
H-GSP
96334
similar in ambulocetids
18470
Remingtonocetid whales
Another main cusp lost, Remingtonocetus IITR-SB 2605
valley gone,
Two large cusps remain,
new cusp added
one new cusp formed
anteriorly.
in front of these.
Andrewsiphius
IITR-SB 2723
Andrewsiphius IITR-SB 3153
Protocetid
Two large cusps remain.
Similar to pakicetids
whales
and ambulocetids.
IITR-SB
3189
IITR-SB
4122
Basilosaurid
Three new cusps
Similar to
whales
added posteriorly.
remingtonocetids,
but with two new cusps
added posteriorly.
MMNS 2339
USNM 11962
figure 34. Left lower (left column) and upper (right column) molars of Eocene whales
and the artiodactyl Indohyus, showing the vast differences in the topography of teeth.
O
utline diagrams for upper molars show how the position of cusps changes in
evolutionary time. Indohyus is discussed in chapter 14.
The Otter Whale | 113
bone: this is called a fused symphysis. An unfused symphysis allows for
some independent movement of the jaws during chewing, and most small
mammals have unfused jaws, for instance dogs, cats, rabbits, and rats.
Larger animals that eat plants, such as horses, cows, elephants, rhinos,
and hippos, tend to have fused symphyses. Large, long-snouted animals,
such as crocodiles, also tend to have fused symphyses. The long symphy-
sis, fused or unfused, gives the jaw strength when it is slapped shut, pos-
sibly preventing teeth from interlocking incorrectly (misocclusion).
Breathing and Swallowing. The nose opening of remingtonocetids is
near the tip of the snout, which is where it is in land mammals. With the
long snout, it may have allowed the whales to lie in wait in deeper water,
with the tip of their nose above the water, and thus avoid the need to
come to the surface to breathe while hunting. However, a nose opening
that far forward might be helpful for other reasons. For remingtonocetids,
freshwater conservation, in the face of the salty ocean they lived in, could
be important. Modern seals use their nasal cavity to retrieve water from
exhalation: water vapor coming from the lungs condenses in the nasal
cavity and is taken up by the tissues inside the nose.12 It is possible that
the long nasal cavity in remingtonocetids served this function.
The remainder of the skull has its peculiarities too. The hard palate
of Remingtonocetus extends nearly to the area of the ears, as in Ambu-
locetus, although it does not reach as far down (ventral) as in Ambu-
locetus. There is a prominent midline crest on the hard palate, probably
for the attachment of the chewing muscles that attach to its lateral side:
the left and right medial pterygoid muscles. The medial pterygoid is a
powerful mouth-closing muscle, useful in a fish eater that clamps its
teeth down fast on prey. In closing the mouth, the medial pterygoid
works with two other muscles: masseter and temporalis. Temporalis
attaches to the side of the skull, as well as to a crest on top of the skull
(the sagittal crest, figure 29), and those bony attachments suggest that it