Sie sind auf Seite 1von 61

There are few scientists of whom it can be said that their mistakes are more int

than their colleagues' successes, but Albert Einstein was one. Few "blunders" ha
ve had a
longer and more eventful life than the cosmological constant, sometimes describe
d as the
most famous fudge factor in the history of science, that Einstein added to his t
heory of
general relativity in 1917. Its role was to provide a repulsive force in order t
o keep the
universe from theoretically collapsing under its own weight. Einstein abandoned
cosmological constant when the universe turned out to be expanding, but in succe
years, the cosmological constant, like Rasputin, has stubbornly refused to die,
itself to the fore, whispering of deep enigmas and mysterious new forces in natu
whenever cosmologists have run into trouble reconciling their observations of th
universe with their theories.
This year the cosmological constant has been propelled back into the news as an
explanation for the widely reported discovery, based on observations of distant
stars, that some kind of "funny energy" is apparently accelerating the expansion
of the
universe. "If the cosmological constant was good enough for Einstein," the cosmo
Michael Turner of the University of Chicago remarked at a meeting in April, "it
be good enough for us."
Einstein has been dead for 43 years. How did he and his 80-year-old fudge factor
come to
be at the center of a revolution in modern cosmology?
The story begins in Vienna with a mystical concept that Einstein called Mach's p
Vienna was the intellectual redoubt of Ernst Mach (1838-1916), a physicist and
philosopher who bestrode European science like a Colossus. The scale by which
supersonic speeds are measured is named for him. His biggest legacy was philosop
he maintained that all knowledge came from the senses, and campaigned relentless
against the introduction of what he considered metaphysical concepts in science,
for example.
1Another was the notion of absolute space, which formed the framework of Newton'
universe. Mach argued that we do not see "space," only the players in it. All ou
knowledge of motion, he pointed out, was only relative to the "fixed stars." In
his books
and papers, he wondered if inertia, the tendency of an object to remain at rest
or in
motion until acted upon by an outside force, was similarly relative and derived
from an interaction with everything else in the universe.
"What would become of the law of inertia if the whole of the heavens began to mo
ve and
stars swarmed in confusion?" he wrote in 1911. "Only in the case of a shattering
of the

universe do we learn that all bodies, each with its share, are of importance in
the law of
Mach never ventured a guess as to how this mysterious interaction would work, bu
Einstein, who admired Mach's incorrigible skepticism, was enamored of what he
sometimes called Mach's principle and sometimes called the relativity of inertia
. He
hoped to incorporate the concept in his new theory of general relativity, which
completed in 1915. That theory describes how matter and energy distort or "curve
" the
geometry of space and time, producing the phenomenon called gravity.
In the language of general relativity, Mach's principle required that the spacetime
curvature should be determined solely by other matter or energy in the universe,
and not
any initial conditions or outside influences -- what physicists call boundary co
Among other things, Einstein took this to mean that it should be impossible to s
olve his
equations for the case of a solitary object -- an atom or a star alone in the un
iverse -since there would be nothing to compare it to or interact with.
So Einstein was surprised a few months after announcing his new theory, when Kar
Schwarzschild, a German astrophysicist serving at the front in World War I, sent
him just
such a solution, which described the gravitational field around a solitary star.
"I would
not have believed that the strict treatment of the point mass problem was so sim
Einstein said.
Perhaps spurred in part by Schwarzschild's results, Einstein turned his energies
in the fall
of 1916 to inventing a universe with boundaries that would prevent a star from e
its neighbors and drifting away into infinite un-Machian loneliness. He worked o
ut his
ideas in a correspondence with a Dutch astronomer, Willem de Sitter, which are t
o be
published this summer by the Princeton University Press in Volume 8 of "The Coll
Papers of Albert Einstein." Like most of his colleagues at the time, Einstein co
2the universe to consist of a cloud of stars, namely the Milky Way, surrounded b
y vast
space. One of his ideas envisioned "distant masses" ringing the outskirts of the
Way like a fence. These masses would somehow curl up space and close it off.
His sparring partner de Sitter scoffed at that, arguing these "supernatural" mas
ses would
not be part of the visible universe. As such, they were no more palatable than N
old idea of absolute space, which was equally invisible and arbitrary.
In desperation and laid up with gall bladder trouble in February of 1917, Einste
in hit on
the idea of a universe without boundaries, in which space had been bent around t
o meet
itself, like the surface of a sphere, by the matter within. "I have committed an

suggestion with respect to gravitation which exposes me to the danger of being c
to the nut house," he confided to a friend.
This got rid of the need for boundaries -- the surface of a sphere has no bounda
ry. Such a
bubble universe would be defined solely by its matter and energy content, as Mac
principles dictated. But there was a new problem; this universe was unstable, th
e bubble
had to be either expanding or contracting. The Milky Way appeared to be neither
expanding nor contracting; its stars did not seem to be going anywhere in partic
Here was where the cosmological constant came in. Einstein made a little mathema
fix to his equations, adding "a cosmological term" that stabilized them and the
Physically, this new term, denoted by the Greek letter lambda, represented some
kind of
long range repulsive force, presumably that kept the cosmos from collapsing unde
r its
own weight.
Admittedly, Einstein acknowledged in his paper, the cosmological constant was "n
justified by our actual knowledge of gravitation," but it did not contradict rel
either. The happy result was a static universe of the type nearly everybody beli
eved they
lived in and in which geometry was strictly determined by matter. "This is the c
ore of the
requirement of the relativity of inertia," Einstein explained to de Sitter. "To
me, as long
as this requirement had not been fulfilled, the goal of general relativity was n
ot yet
completely achieved. This only came about with the lambda term."
The joke, of course, is that Einstein did not need a static universe to have a M
achian one.
Michel Janssen, a Boston University physicist and Einstein scholar, pointed out,
"Einstein needed the constant not because of his philosophical predilections but
of his prejudice that the universe is static."
3Moreover, in seeking to save the universe for Mach, Einstein had destroyed Mach
principle. "The cosmological term is radically anti-Machian, in the sense that i
t ascribes
intrinsic properties (energy and pressure-density) to pure space, in the absence
of matter,"
said Frank Wilczek, a theorist at the Institute for Advanced Study in Princeton.
In any event, Einstein's new universe soon fell apart. In another 10 years the a
Edwin Hubble in California was showing that mysterious spiral nebulae were galax
ies far
far away and getting farther -- in short that the universe might be expanding.
De Sitter further confounded Einstein by coming up with his own solution to Eins
equations that described a universe that had no matter in it at all.
"It would be unsatisfactory, in my opinion," Einstein grumbled, "if a world with
matter were possible."
De Sitter's empty universe was also supposed to be static, but that too proved t

o be an
illusion. Calculations showed that when test particles were inserted into it, th
ey flew
away from each other. That was the last straw for Einstein. "If there is no quas
world," he said in 1922, "then away with the cosmological term."
In 1931, after a trip to the Mount Wilson observatory in Pasadena, Calif., to me
et Hubble,
Einstein turned his back on the cosmological constant for good, calling it "theo
unsatisfactory anyway."
He never mentioned it again.
In the meantime, the equations for an expanding universe had been independently
discovered by Aleksandr Friedmann, a young Russian theorist, and by the Abbe Geo
Lemaitre, a Belgian cleric and physicist. A year after his visit with Hubble, Ei
threw his weight, along with de Sitter, behind an expanding universe without a
cosmological constant.
But the cosmological constant lived on in the imagination of Lemaitre, who found
that by
judicious application of lambda he could construct universes that started out ex
slowly and then sped up, universes that started out fast and then slowed down, o
r one that
even began expanding, paused, and then resumed again.
This last model beckoned briefly to some astronomers in the early 1950's, when
measurements of the cosmic expansion embarrassingly suggested that the universe
4only two billion years old -- younger Earth. A group of astronomers visited Ein
stein in
Princeton and suggested that resuscitating the cosmological constant could resol
ve the
age discrepancy. Einstein turned them down, saying that the introduction of the
cosmological constant had been the biggest blunder of his life. George Gamow, on
e of
the astronomers, reported the remark in his autobiography, "My World Line," and
became part of the Einstein legend.
Einstein died three years later. In the years after his death, quantum mechanics
, the
strange set of rules that describe nature on the subatomic level (and Einstein's
bete noire)
transformed the cosmological constant and showed just how prescient Einstein had
in inventing it. The famous (and mystical in its own right) uncertainty principl
e decreed
that there is no such thing as nothing, and even empty space can be thought of a
s foaming
with energy.
The effects of this vacuum energy on atoms had been detected in the laboratory,
as early
as 1948, but no one thought to investigate its influence on the universe as a wh
ole until
1967, when a new crisis, an apparent proliferation of too-many quasars when the
was about one-third its present size, led to renewed muttering about the cosmolo
constant. Jakob Zeldovich, a legendary Russian theorist who was a genius at marr

microphysics to the universe, realized that this quantum vacuum energy would ent
er into
Einstein's equations exactly the same as the old cosmological constant.
The problem was that a naive straightforward calculation of these quantum fluctu
suggested that the vacuum energy in the universe should be about 118 orders of
magnitude (10 followed by 117 zeros) denser than the matter. In which case the
cosmological constant would either have crumpled the universe into a black hole
in the
first instant of its existence or immediately blown the cosmos so far apart that
not even
atoms would ever have formed. The fact that the universe had been sedately and h
expanding for 10 billion years or so, however, meant that any cosmological const
ant, if it
existed at all, was modest.
Even making the most optimistic assumptions, Dr. Zeldovich still could not make
predicted cosmological constant to come out to be less than a billion times the
Ever since then, many particle theorists have simply assumed that for some as-ye
tunknown reason the cosmological constant is zero. In the era of superstrings and
ambitious theories of everything tracing history back to the first micro-micro s
econd of
unrecorded time, the cosmological constant has been a trapdoor in the basement o
5physics, suggesting that at some fundamental level something is being missed ab
out the
world. In an article in Reviews of Modern Physics in 1989, Steven Weinberg of th
University of Texas referred to the cosmological constant as "a veritable crisis
," whose
solution would have a wide impact on physics and astronomy.
Things got even more interesting in the 1970's with the advent of the current cr
op of
particle physics theories, which feature a shadowy entity known as the Higgs fie
ld, which
permeates space and gives elementary particles their properties. Physicists pres
ume that
the energy density of the Higgs field today is zero, but in the past, when the u
niverse was
hotter, the Higgs energy could have been enormous and dominated the dynamics of
universe. In fact, speculation that such an episode occurred a fraction of a sec
ond after
the Big Bang, inflating the wrinkles out of the primeval chaos -- what Dr. Turne
r calls
vacuum energy put to a good use -- has dominated cosmology in the last 15 years.
"We want to explain why the effective cosmological constant is small now, not wh
y it
was always small," Dr. Weinberg wrote in his review. In their efforts to provide
explanation, theorists have been driven recently to talk about multiple universe
connected by space-time tunnels called wormholes, among other things.
The flavor of the crisis was best expressed, some years ago at an astrophysics c
by Dr. Wilczek. Summing up the discussions at the end of the meeting, he came at

last to
the cosmological constant. "Whereof one cannot speak, thereof one must be silent
," he
said, quoting from Ludwig Wittgenstein's "Tractatus Logico-Philosophicus."
Now it seems that the astronomers have broken that silence.
Copyright 2002 The New York Times Company
6Mysteries of the Universe
Quantum Theory Tugged, and All of Physics Unraveled
They tried to talk Max Planck out of becoming a physicist, on the grounds that h
ere was
nothing left to discover. The young Planck didn't mind. A conservative youth fro
m the
south of Germany, a descendant of church rectors and professors, he was happy to
add to
the perfection of what was already known.
Instead, he destroyed it, by discovering what was in effect a loose thread that
tugged would eventually unravel the entire fabric of what had passed for reality
As a new professor at the University of Berlin, Planck embarked in the fall of 1
900 on a
mundane sounding calculation of the spectral characteristics of the glow from a
object. Physicists had good reason to think the answer would elucidate the relat
between light and matter as well as give German industry a leg up in the electri
c light
business. But the calculation had been plagued with difficulties.
Planck succeeded in finding the right formula, but at a cost, as he reported to
the German
Physical Society on Dec. 14. In what he called "an act of desperation," he had t
o assume
that atoms could only emit energy in discrete amounts that he later called quant
a (from
the Latin quantus for "how much" ) rather than in the continuous waves prescribe
d by
electromagnetic theory. Nature seemed to be acting like a fussy bank teller who
not make change, and would not accept it either.
That was the first shot in a revolution. Within a quarter of a century, the comm
on sense
laws of science had been overthrown. In their place was a bizarre set of rules k
nown as
quantum mechanics, in which causes were not guaranteed to be linked to effects;
subatomic particle like an electron could be in two places at once, everywhere o
nowhere until someone measured it; and light could be a wave or a particle.
and the unknown, who as been accorded some of the ultimate accolades in pop cult
ure -appearing as Einstein's poker buddy on "Star Trek: The Next Generation," and as
a guest
star on "The Simpsons."
While a graduate student, in 1963, he learned he had amyotrophic lateral scleros
is and
was given a few years to live. He has moved about in a wheelchair for more than
25 years

and now speaks only through a voice synthesizer. Dr. Hawking, for whom the word
"puckish" seems to have been invented, has often said his disability is an advan
because it frees him to sit and think. Next month his colleagues will celebrate
his 60th
birthday with a weeklong all-star symposium in Cambridge.
In the new book's introduction, Dr. Hawking admits that "A Brief History of Time
" was
"not easy going" and laments that some readers got stuck and did not finish it.
He has
tried, he says, to make this one easier. Slightly longer than the earlier book,
"Nutshell," at
216 pages, is embellished with colorful illustrations that give it a coffee-tabl
e-book look.
So far the critics are in qualified agreement; one, Bryan Appleyard in the The N
he Bivalvia, the second largest class within the
Solnhofen Limestone of Eichsta tt, Germany, and was
described by Cosimo Collini (17271806) in 1784.
Collini concluded that it was a possible sea creature
of unknown affinity, although he did note bat-like
features. In 1801, the great French anatomist Georges
Cuvier (17691832) recognized that the creature
was a reptile and that its elongated digits must have
supported flight membranes. Cuvier was thus the first
to recognize pterosaurs as flying reptiles and, in 1809,
he coined the name Ptero-Dactyle. This later became
the generic name Pterodactylus (Figures 1 and 4).
In the decades that followed, a succession of further
pterosaurs from the Solnhofen Limestone was announced, many in a spectacular state of preservation
and some with their wing membranes intact. The
first recognized British pterosaur, a specimen of the
deep-skulled Dimorphodon, was discovered by Mary
Anning (17991847) in 1827 in Lower Jurassic rocks
of Lyme Regis, Dorset. We now know that Gideon
Mantell (17901852), best known for the discovery
of Iguanodon, found pterosaur remains before this
in the Early Cretaceous Wealden strata of Sussex,
but had thought that these were from birds. North
America yielded its first pterosaur to the prolific
palaeontologist O. C. Marsh (18311899) in 1871
and, by 1876, Marsh had recognized it as a new,
distinctive genus he named Pteranodon (meaning
winged and toothless). With an estimated wingspan
of 6 m, Pteranodon was huge compared to most
earlier discoveries.
While these discoveries and others were being
made, varied opinions on the nature and life style of
pterosaurs were appearing, and they were variously
depicted as swimming creatures, as bats, marsupials,
or as kin of birds. By the early 1900s, it was generally
agreed that pterosaurs were bat-like flying reptiles and,
in 1901, Harry Seeley (18391909) published Dragons
of the Air, the first book devoted to pterosaurs.
South American Cretaceous pterosaurs have proved
to be among the most important in the world, but not
until 1971 was the first pterosaur from the now famous
Santana Formation of Brazil discovered. Since then a
significant number of new kinds from around the

world (around 70 genera are presently recognized)

have revealed previously unimagined morphologies
and maximum sizes. Until 1971, Pteranodon sternbergi
Figure 1 Life restoration of the Late Jurassic pterodactyloid, Pterodactylus , f
rom the German Solnhofen Limestone in a quadrupedal
stance. Note the presence of body hair and the soft tissue head crest. Reproduce
d with permission from Dino Frey. Buffetaut E
and Mazin J M (2003) Evolution and Palaeobiology of Pterosaurs. Geological Socie
ty Special Publication 217 . London: The Geological Society
of London.510 FOSSIL VERTEBRATES/Flying Reptiles
(wingspan, 9 m) was the largest known flying animal,
but the discovery in Texas of Quetzalcoatlus revealed
that the biggest pterosaurs achieved wingspans of
11 m. Related pterosaurs of similar or larger size were
discovered in the 1990s in Spain and eastern Europe.
The Pterosaur Skeleton
The pterosaur skeleton was highly modified for flight,
and the most obvious features are the huge size of the
skull compared with the body and the extreme
elongation of one of the fingers. Like birds, most
pterosaurs had hollow bones with foramina (small
openings), indicating that they contained air sacs
connected to the lungs. Pterosaur bones were supported internally by struts, and the bone walls themselves, usually no thicker than 2 mm, are composed
of multiple overlapping layers and thus combine
lightness with strength.
Pterosaur skull morphology is varied, although the
majority had long, slim, shallow jaws and all had
large orbits (eye-sockets). In basal pterosaurs, the
external nostril was separate from an opening in
front of the orbit called the antorbital fenestra. In
pterodactyloids, these two openings merged into a
single one called the nasoantorbital fenestra. Pterosaur teeth were extremely variable. Widely spaced
pointed teeth, were widespread and from ancestors
with teeth like these evolved species with fang-like
teeth at the jaw tips and the unique Istiodactylus
with its short petal-shaped teeth. The Late Triassic
Eudimorphodon and Austriadactylus possessed
multicusped teeth while elongate, slender teeth
numbering in the hundreds evolved in the ctenochasmatoids. Toothlessness evolved several times. Some
pterosaurs skulls sport bony crests at the jaw tips,
along the midline or at the back of the skull.
Figure 2 Wing skeleton of an ornithocheiroid pterodactyloid.
Unlike birds and bats, the main wing spar in pterosaurs was formed by a hypertrophied digit (Figure 2).
This wing finger is generally considered to be the
fourth because the digital formulae of the pterosaur
hand best matches that of digits one to four in the
hands of other reptiles. However, a rod-shaped bone
projecting from the pterosaur wrist, called the pteroid
bone, has at times been argued to represent the first
hand digit. This is a minority view today but, if it is
correct, then pterosaurs have five hand digits and the
wing finger is the fifth. Although most pterosaur
fossils show the pteroid pointing towards the shoulder, some workers suggest that it pointed forwards
parallel to the neck. Regardless, the pteroid was

probably mobile and used to control the attitude of

the propatagium (Figure 3).
The pterosaur pectoral girdle includes a (normally
fused) scapula and coracoid that meet at an acute
angle and, as expected for flying animals, the socket
for the humerus faces sideways and slightly upwards.
The coracoids attach to an enlarged keeled sternum
that anchored most of the major flight muscles. The
bones of the pterosaur pelvis were short, usually fused
together, and with a closed hip socket. Pterosaurs
have a pair of unique rod- or plate-like bones called
the prepubes projecting forwards from the bottom
of the pelvis. Like the gastralia (belly ribs) that all
pterosaurs possessed, they may have helped support
the gut or keep the abdomen rigid.
The vertebral column in pterosaurs can clearly be
differentiated into cervical, dorsal, sacral, and caudal
portions. The number of vertebrae is variable and
pterosaurs have 79 cervical, 1116 dorsal, 310
sacral, and 1140 caudal vertebrae. The many caudal
vertebrae of basal pterosaurs are encased in long bony
processes that make the tail stiff and rod-like. In
derived Cretaceous pterosaurs, most of the dorsalFOSSIL VERTEBRATES/Flying Repti
les 511
vertebrae are fused together forming a structure
called the notarium.
Compared with the other wing segments, the pterosaur humerus is short, although generally with massive crests for muscle attachment. The ulna is always
larger than the radius and both are attached distally
to block-shaped carpal bones. Projecting from one
of these is the unique pteroid. Pterosaurs had four
metacarpals, the first three of which were slim and,
with the exception of Nyctosaurus from Late Cretaceous North America, attached to short, clawed
fingers. Why Nyctosaurus lacked clawed fingers is
unknown, but in all other pterosaurs these digits
may have served important functions. Trackways
show that they were used in walking, and it is also
possible that they were employed in grooming or
climbing. The fourth metacarpal was robust and
tipped with a twisted, roller-like distal end to which
was attached the massive wing finger. This consists of
four long straight bones, excepting a few genera
where there were only three. Because of the twisted
end of the fourth metacarpal, the wing finger would
have lain parallel to the bodys long axis when the
wing was folded up.
The pterosaur hind limb is lightly built and the
head of the femur is only slightly offset from the
long axis of the shaft. Pterosaur hind limbs seem to
have been quite flexible, but mostly sprawled to the
sides. During flight, the hind limb was probably held
in a bat-like orientation and could have been used to
control the shape of the wing membranes. Basal
pterosaurs are five-toed, with a prominent curving
Figure 3 Schematic representation of the flight membranes in
a generalized pterodactyloid pterosaur.
fifth digit that is hooked towards the tail. In pterodactyloids, the fifth digit is either absent or present as

a tiny stub.
Because some articulated fossils indicate that the
foot could assume a 90  angle relative to the tibia
(and there is little evidence for much motion at the
metatarsophalangeal joints), pterosaurs have generally been regarded as plantigrade (placing the whole
length of the foot on the ground when walking). In
1983, Kevin Padian argued that this was not the case
for Dimorphodon and that it may instead have been
digitigrade (walking only on the toes). This was later
inferred for all pterosaurs. An articulated Dimorphodon foot shows, however, that only limited motion
was possible at the metatarsophalangeal joint, thus
supporting a plantigrade posture. This is in agreement
with probable pterosaur tracks preserved as trace
Soft Tissue, Integument, and
Pterosaur Life Appearance
Many aspects of pterosaur life appearance remain
unknown or controversial, although a number of exceptional fossils have provided some surprising
details. Pterosaur body hair was reported as early as
1831 and described for various Jurassic pterosaurs
between the 1920s and 1970s and today it is clear
that pterosaurs had bristle-like hairs covering their
necks and bodies (Figure 4). The active flapping flight
and body hair of pterosaurs suggest that they had an
elevated metabolism.
Other exceptional fossils show that some pterosaurs possessed a throat pouch, webbing between
the toes, and scales on the soles of the feet. Soft
Figure 4 An exceptionally well preserved skeleton of the Late
Jurassic pterodactyloid Pterodactylus from the German Solnhofen
Limestone. This specimen preserves parts of the flight mem
branes, a throat pouch, and hairs on the neck and back.512 FOSSIL VERTEBRATES/Fl
ying Reptiles
Figure 5 Variation in skull crest morphology in pterodactyloids. Soft tissue cre
sts are now known for a wide diversity of
pterodactyloids. Reproduced with permission from Dino Frey and Marie Celine Buch
y. Buffetaut E and Mazin J M (2003) Evolution
and Palaeobiology of Pterosaurs. Geological Society Special Publication 217 . Lo
ndon: The Geological Society of London.
tissue skull crests connected to the underlying bony
crests have proved to be widespread and appear
to have doubled the size of the bony crests (Figures 5
and 6). An unexpected discovery is a soft tissue crest
in Pterodactylus, a genus that lacks a bony crest
(Figure 1). The presence of a distinctive bone texture
on the pterosaur snout, jaw, and palate indicates that
pterosaurs were beaked.
Pterosaur wing membranes are known from wellpreserved specimens from the Solnhofen Limestone
and the Early Cretaceous Brazilian Crato and Santana formations. A membrane called the propatagium
extended from the shoulder to the pteroid and perhaps distally to encompass the first three fingers. The
main flight membrane, the brachiopatagium (also
called the cheiropatagium), extended from the tip of
the wing finger to the hind limb, extending
as far distally as the knee, shin, or ankle. Another

membrane, the uropatagium, was present between

the hind limbs (Figure 3). The wing membrane
appears to have been complex, with a thin epidermis,
a layer of vascular tissues, a layer of stiffening fibres
called aktinofibrils, a thin sheet of muscle, and a
Figure 6 Skull of the tapejarid pterosaur Topejara navigaus from
the Early Cretaceous Crato Formation of Brazil with bony and soft
tissue skull crest. Buffetaut E and Mazin J M (2003) Evolution and
Palaeobiology of Pterosaurs. Geological Society Special Publication 217 .
London: The Geological Society of London.FOSSIL VERTEBRATES/Flying Reptiles 513
blood capillary network. Rhamphorhynchus and
probably other long-tailed pterosaurs possessed a
vertical diamond-shaped membrane at the tail tip.
With skin membranes connecting the wings, body,
and legs, pterosaurs may have been superficially batlike but, because bats are mostly dark-coloured nocturnal animals, it is doubtful that the similarities were
strong. Pterosaurs mostly seem to have been ecological
analogues of sea- and water-birds, and it might be
that they were patterned in whites, blacks, and greys,
although bright colours presumably decorated their
pterosaur hind limbs are only superficially similar to
those of dinosaurs, and that re-analysis favoured a
position for pterosaurs outside of the crocodilian
dinosaur group. Rather more heterodox recent ideas
include the suggestion that pterosaurs are the closest
relatives of birds and that pterosaurs are part of the
Several different models have been proposed for
the origin of pterosaurs, but the presence in basal pterosaurs of climbing features and of various details in the
hind limb and pelvis indicative of a leaping ability suggest that pterosaurs first evolved as tree-climbing
The Affinities and Origin of Pterosaurs
Historically, pterosaurs have been allied with Mesozoic marine reptiles, bats, marsupials, and birds
(see Fossil Vertebrates: Dinosaurs; Birds). However,
major improvements in the understanding of vertebrate evolution allowed the palaeontologists of the
nineteenth and twentieth centuries to realize that
pterosaurs were related at least vaguely to dinosaurs
(see ) and their allies.
Although it is clear that pterosaurs are part of the
major reptile assemblage known as the Diapsida,
their affinities within this group are controversial.
The presence of an antorbital fenestra has conventionally meant that pterosaurs have been regarded
as archosaurs, the so-called ruling reptile group
that incorporates crocodilians, dinosaurs, and kin.
Among archosaurs, pterosaurs share a simple hingelike ankle joint with dinosaurs and consequently
have been regarded as close relatives of dinosaurs in
most studies. This view was developed at a time when
some workers thought that pterosaurs originated
from terrestrial bipedal ancestors and that pterosaurs
themselves were bipedal and digitigrade. A small bipedal, long-legged archosaur from Late Triassic
Scotland, Scleromochlus, was argued to be a ptero-

saur ancestor, but recent studies refute this idea. The

idea that pterosaurs might be close relatives of dinosaurs can certainly be regarded as the mainstream
view in vertebrate palaeontology today. However,
several recent studies have questioned the evidence
for this proposed affinity.
An alternative hypothesis argues that pterosaurs
belong instead to a group of archosaur-like diapsids,
the Prolacertiformes. Most prolacertiforms were
superficially lizard-like, but Sharovipteryx from Late
Triassic Kyrgyzstan appears to be intermediate between conventional prolacertiforms and pterosaurs.
It has pterosaur-like hind limbs and vertebrae and
membranes between its hind limbs and tail.
Some other models for pterosaur ancestry have been
proposed. In 1996, S. Christopher Bennett argued that
Pterosaur Diversity and Phylogeny
It was recognized in 1901 that pterosaurs could be
divided into two groups: the toothed, mostly longtailed Rhamphorhynchoidea, and the short-tailed
Pterodactyloidea (including both toothed and toothless kinds). Today, it is clear that rhamphorhynchoids
include the ancestors of pterodactyloids and, consequently, Rhamphorhynchoidea is a grade and not a
clade. Pterosaurs previously referred to as rhamphorhynchoids are nowadays termed basal pterosaurs or
non-pterodactyloids. Although basal pterosaurs were
diverse, it is notable that they were small compared
with the majority of Cretaceous pterodactyloids.
The evolutionary relationships of pterosaurs are
relatively understudied and only recently has pterosaur
phylogeny been analysed. Although some areas of consensus have emerged, authors disagree on the details.
We follow the phylogeny proposed by David Unwin
of the Museum fu r Naturkunde in Berlin (Figure 7).
Perhaps the most basal pterosaur is Preondactylus
from the Late Triassic of Italy. This form has a shorter
coracoid and humerus and longer legs than other
pterosaurs. Dimorphodontids, which include Dimorphodon from Early (and perhaps Middle) Jurassic
Figure 7 Cladogram depicting the relationships of all the major
pterosaur groups. Reproduced from David Unwin.514 FOSSIL VERTEBRATES/Flying Rept
England and Mexico, are basal pterosaurs with deep
skulls superficially like those of puffins, whilst anurognathids were unusual in having short, broad snouts
and abbreviated tails. A surprising recent discovery
is the persistence of anurognathids into the Early
Cretaceous. Two basal pterosaurs, Eudimorphodon
from Late Triassic Italy and Greenland and Campylognathoides from Early Jurassic Germany and India,
are united in the Campylognathoididae based on a
distinctive lower jaw in which two pairs of large
conical teeth are followed by multiple smaller ones.
Rhamphorhynchids were successful Jurassic pterosaurs known from Eurasia, North America, and
Africa. Rhamphorhynchus from Late Jurassic Europe
exhibits a laterally compressed, ventrally directed
lower jaw tip and an array of forward-pointing
teeth. It probably used these to grab fish and other

small prey from the water. Another rhamphorhynchid

lineage, the scaphognathines, had deeper skulls with
teeth perpendicular to the jaw margins.
The Pterodactyloids
Pterodactyloids, the advanced short-tailed pterosaurs,
originated from a rhamphorhynchid-like ancestor
during the Middle Jurassic. The pterodactyloid radiation consisted of four major groups: the robust-jawed
ornithocheiroids, the slim-jawed ctenochasmatoids,
the low-crested dsungaripteroids, and the long-necked,
crested azhdarchoids. A fifth group, the lonchodectids
from Early Cretaceous England, are of uncertain affinity. Lonchodectids were small (wingspan, 12 m) with
long, dorsoventrally flattened jaws with small teeth,
each of which was supported by a low bony collar at
its base.
Ornithocheiroids were large predatory pterosaurs
(wingspan, 29 m) with robust beaks, often housing
recurved, fang-like teeth at their tips (Figure 8). Their
jaws frequently possessed keel-like dorsal and ventral
crests, and some forms also possessed crests on the
back of the skull. The toothless pteranodontids and
nyctosaurids appear to be members of this group. The
earliest known ornithocheiroids appear at the start of
the Cretaceous, while nyctosaurids survived to the
very end of the Cretaceous.
Ctenochasmatoids had needle-like meshes of teeth
set in long, thin jaws. In Pterodaustro from late
Early Cretaceous Argentina, the upturned lower jaw
contains approximately 1000 bristle-like teeth. These
were surely used for filtering small organisms from
the water. Unlike ornithocheiroids, ctenochasmatoids
had elongate cervical vertebrae and were generally
small (wingspan, 50 cm to 2 m), although Cearadactylus from Early Cretaceous Brazil was a giant with
a wingspan of 5.5 m.
Figure 8 Jaw tips of the Cretaceous ornithocheiroid pterosaur
Coloborhynchus . Note the massive fang like anterior teeth and the
low keel like crests on both jaws. This specimen is from the
Santana Formation of Brazil and would have belonged to an
animal with a skull of approximately 1 m in length.
Members of the Dsungaripteroidea are known from
the Late Jurassic and Early Cretaceous of Eurasia,
Africa, and South America. Like some ctenochasmatoids, dsungaripteroids had a midline crest on the top of
the skull. Their beaks often had toothless tips and they
may have been predators of molluscs and crustaceans.
Finally, the strangest pterodactyloids must be the
azhdarchoids. These include the long-necked azhdarchids and the crested tapejarids. Azhdarchids
may have exceeded wingspans of 11 m and were
widely distributed in the Late Cretaceous. They may
have been ecological generalists akin to storks, and
were probably not specialist carrion feeders or mud
probers as has been proposed. Determining the life
style of the tapejarids is more difficult. The vaguely
parrot-like skull of Tapejara (Figure 6) from Lower
Cretaceous Brazil led some workers to propose that
it was a fruit eater, but it might better be imagined
as an auk analogue. Recently, it has been suggested

that Thalassodromeus, also from Lower Cretaceous

Brazil, was a fish eater that trawled its blade-like
lower jaw through the water.
Pterosaur Palaeobiology
Because pterosaurs are unique and extinct, reconstructing their palaeobiology is difficult and nothing
is known about several aspects of their lives. Limited
evidence does allow us, however, to reconstruct their
sensory abilities, feeding behaviours, and styles of
The large orbits of pterosaurs show that they had
large eyes, and the abundance of visual display
phylum Mollusca, is one of the most familiar of all
invertebrate taxa. Modern representatives, such as
mussels, cockles, oysters, and scallops, are well
known from excursions to the coast, and in many
parts of the world they are important commercial
species. Their generally excellent fossil record has
allowed their evolutionary history to be traced back
to the Early Palaeozoic and, for much of this time,
they have been important components of many
faunas. From rather modest beginnings, they have
conquered a range of habitats from the deep sea to
freshwater, exploited a wide range of life habits
(from deep burrowing to swimming), and undergone
a near-exponential taxonomic proliferation, a spectacular example of an adaptive radiation.
Bivalves come in all manner of shapes and sizes,
from tiny, thin-shelled commensals that live in association with sea anemones, to giant clams and the
extinct rudists and inoceramids which reach(ed)
sizes well over 1 m. Shell morphology is extremely
plastic, but all are modifications of the same basic
theme. The intimacy of the shell morphology to life
habit has been a great benefit in reconstructing the life
habits of extinct bivalves, but has also frustrated
many attempts to establish the relationships between
different groups within the class. Bivalves have been
proven to be good palaeoenvironmental indicators,
but they have only limited use in biostratigraphy.
Freshwater mussels have been used to date fluvial
deposits in the Carboniferous Coal Measures of
Western Europe, and inoceramids have been used
for Late Cretaceous deep marine settings (e.g., in
New Zealand). In general, however, species of the370 FOSSIL INVERTEBRATES/Bivalv
class are too long lived and too facies specific to be of
any great value.
General Morphology
As the name implies, bivalves comprise two calcareous valves. These are arranged laterally (left and
right), are joined dorsally by a partially calcified elastic ligament, and enclose the soft tissue. Each valve
has clearly differentiated posterior and anterior features, i.e., inequilateral. The primitive arrangement,
retained by most bivalves, was to have a plane of
symmetry parallel to the commissure (the join between the two valves), resulting in valves which are
mirror images of one another (i.e., equivalve). Although this symmetry is found in virtually all bivalves

which live with the commissural valve perpendicular

to the substrate surface (orthothetic), it has been lost
in those which have adopted a pleurothetic habit
where they lie on one valve (e.g., oysters, scallops).
In these cases, there is a tendency for the two valves to
become dissimilar (i.e., inequivalve), typically with
the underlying valve becoming more bowl-like and
the upper one more reduced like a lid.
Shell Morphology
All bivalves possess a pair of shells which may be
shut to provide protection from both environmental
stresses (e.g., desiccation in the intertidal habitat) and
the threat of predation. Most shells are reasonably
robust, which has provided the class with a generally
excellent fossil record. Although shell morphology in
bivalves is very variable and intimately linked to their
life habits (see below), all shells are simple modifications of the basic shell secretion model used by all
shelled molluscs. The shell is secreted by the mantle
lobes and grows by marginal accretion, as evidenced
by the growth lines on the surface of the valve
(Figures 1A and 2A). These growth lines are particularly marked in bivalves from intertidal and shallow
temperate habitats, where the animals experience
pronounced seasonality and largely stop growing
during the winter months. Bivalves which experience
more equable conditions do not show such obvious or
regular patterns. Inspection of the growth lines in
sectioned valves shows that, although most shell material is added ventrally, the shell is also thickened
during growth (Figure 2B), demonstrating that the
entire mantle surface is responsible for adding material. The outermost part of the shell is an organic layer
called the periostracum secreted at the mantle edge
(Figure 2C). The thickness of the periostracum varies
between taxa, from less than 1 mm in oysters and
Figure 1 (A) Mercenaria mercenaria , a shallow burrowing bi
valve from the Pliocene of Florida. Note the prominent annual
growth bands. (B) Pecten maxima , a free living epifaunal scallop
from the Holocene Atlantic.
scallops to several hundred micrometres in some
mussels. In many cases, the periostracum is lost by
abrasion and decay during the life of the animal,
particularly on the older parts of the shell, and there
is no real prospect of it being preserved in any but the
most exceptional circumstances. The primary function of the periostracum is to act as the template on
which the calcareous part of the shell is deposited, but
it may also provide protection from both corrosive
waters and predators that dissolve the shell. It is
particularly noticeable that freshwater bivalves have
very thick periostraca.
The main part of the shell, however, is calcareous.
It is in effect a ceramic made up of calcium carbonate
crystals in an organic matrix (the latter accounting for
<5% of the dry weight of the shell). The proteinaceous matrix controls both the polymorph of calcium carbonate used and the arrangement of the
crystals. All bivalves contain aragonite in their shells
and the vast majority are wholly so. Some taxa, how-

ever, chiefly those exploiting epifaunal life habits, also

secrete calcite in their outer layers. The oysters have
taken this to its extreme and the bulk of the shell isFOSSIL INVERTEBRATES/Bival
ves 371
Figure 2 Details of shell formation in a generalized bivalve.
(A) Marked comarginal growth lines on the shell surface.
(B) Section through the shell along the line indicated in (A) show
ing the arrangement of growth lines within the shell. (C) The
relationship between the shell and the underlying mantle edge
(close up details of the circled area in (B)). i, inner mantle fold; m,
middle mantle fold; o, outer mantle fold.
calcitic, with aragonite being confined to the sites of
muscle attachment and the ligament.
Molluscan shell is immensely strong, in fact often
much stronger than vertebrate bone. There are a
number of different microstructures (Figure 3), each
with different mechanical properties, and most shells
are made up of two or three arranged in different
layers. Different taxa show different arrangements
and these are considered to be of phylogenetic significance. It is apparent that the earliest bivalves were
wholly aragonitic and chiefly composed of nacre
(Figure 3A), and that subsequent evolution has produced the wide array of microstructural arrangements
seen today. The effect of differing crystal sizes, amount
of organic material, and polymorph used has affected
the preservation potential of different taxa; many of
the Palaeozoic and Mesozoic taxa that were originally
aragonitic are either preserved as internal moulds or
are replaced by calcite.
Details of the internal features of the shell are
shown in Figure 4. The hinge plate is situated dorsally
and houses the ligament and teeth. The ligament is an
elastic, partially calcified layer that provides a very
energy-efficient opening mechanism. During valve
closure, energy is stored in the ligament as it is flexed
by the contracted adductor muscle(s) (Figure 4C).
When the muscle is relaxed, the ligament springs the
valves apart causing them to gape. This passive valve
opening mechanism is the reason why many fossil
bivalves are found in a disarticulated state. Although
the ligament itself is seldom preserved, its position
may be inferred from the presence of the ligament
pits in which it is anchored (Figures 4 and 5). Most
bivalves have teeth on the hinge plate which fit into
corresponding sockets on the opposite valve and function to keep the valves in perfect alignment. Both
ligamenture and dentition vary markedly amongst
higher taxa of bivalves, and both are often used as
informative characters in establishing phylogenies.
Some of the range of hinge plate architectures is
shown in Figure 5.
A number of attachment scars mark the locations
where muscles are anchored to the shell. The most
significant of these are the adductor scars (Figures 4A
and 4B). If the adductor scars are paired (i.e., dimyarian), they occur posteriorly and anteriorly. If an
animal is monomyarian, the single muscle (the posterior) occupies a more central position. In many
taxa, there is a thin pallial line running around the

shell a small distance from the ventral edges that

marks the attachment of the mantle to the shell. In
infaunal taxa, where the posterior mantle has been
fused and elongated to form siphons, the pallial line is
inflected forming the pallial sinus. The sinus represents the space into which the siphons are withdrawn
when the valves are shut. Other muscle attachment
scars may be more or less apparent, including the
insertions of the pedal musculature (particularly in
burrowers and byssate taxa).
Soft Part Anatomy
Bivalves are laterally compressed and, unlike most
molluscs, there is no head or radula. The internal
organs are enclosed by the two mantle lobes that are
joined dorsally (Figure 4C). The chief function of the
mantle is to secrete the shells, but the ventral edges of
each mantle lobe are differentiated into three folds
(Figure 2C), only the outermost of which is directly
concerned with shell manufacture. The innermost
fold controls water flow into and out of the mantle
cavity, whilst the middle fold has sensory capability.
In several bivalve groups (such as scallops), the
middle fold is well developed with tentacles and
eyes. In some taxa, the mantle is extended posteriorly
and fused to form a pair of siphons through which
water is directed into (inhalant) and out of (exhalant)
the mantle cavity.372 FOSSIL INVERTEBRATES/Bivalves
Figure 3 Scanning electron micrographs of four of the most common bivalve shell
microstructures. (A) Aragonitic nacre (from the
inner shell layer of Pinna nobilis ). (B) Aragonitic crossed lamellar structure
(from the outer layers of Corbula gibba ). (C) Foliated calcite
(from Ostrea edulis ). (D) Calcitic prisms (from the outer shell layer of Pinna
nobilis ). Scale bars for (A) (C) represent 10 mm and for
(D) represents 100 mm.
The mantle cavity is spanned by one or two adductor
muscles that attach to the shell and act antagonistically
with the ligament to close the valves on contraction
(Figure 4C). A significant part of the mantle cavity is
occupied by a pair of gills (the ctenidia) lying on either
side of the rest of the viscera. In most bivalves, the gills
are involved with both respiration and ciliary suspension feeding (filtering small particles out of the water
which are then transferred to the mouth by a pair of
labial palps). Recent bivalves show a number of different gill morphologies depending largely on the feeding process employed. Deposit feeders, e.g., Nucula,
have less well-developed (protobranch) gills, whilst
members of the Lucinidae augment their filter feeding
by energy gained from the activities of sulphide-oxidizing chemosymbiotic bacteria living within the modified
gills. The carnivorous septibranch bivalves (e.g.,
Cuspidaria, Poromya) suck in small prey (such as
amphipods) using their modified siphons. These extraordinary bivalves have lost their gills and respire
over the inner surface of the mantle. Other significant
organs within the mantle cavity include the gut, heart,
circulatory system, and the foot. The gut runs between
the anteriorly positioned mouth and the posterior anus,
and includes a complex stomach which, again, has a
number of configurations depending on the feeding

biology of the animal. Blood is circulated throughout

the animal by a three-chambered heart. A muscular
foot is present in all juvenile and most adult bivalves
and occupies the centre of the mantle cavity.
Naturally, the soft part anatomy of bivalves is very
seldom preserved, although preservation of gill and
muscle material has been reported in exceptional circumstances. Various details, however, can be inferred
from the study of the internal surface of the shells.
Apart from adductor muscle scars, the practised eye
may pick out the attachment points of more minor
muscles and impressions of radial muscles and blood
vessels within the mantle.
Modern bivalves exploit a wide range of life habits.
Many burrow to varying depths within soft sediments,FOSSIL INVERTEBRATES/Bivalve
s 373
Figure 4 Various aspects of the internal morphology of bivalve
shells. (A) The left valve of a generalized burrowing dimyarian.
(B) The right valve of a generalized byssate scallop. (C) The rela
tionship between the two valves and their associated mantle
lobes, adductor musculature and ligament.
but others attach to or bore into hard substrates by a
variety of means; others have become free living, some
with the ability to swim. Most bivalves are marine,
exploiting niches from the intertidal zone down into
the abyssal depths, but successful groups (including
the oysters) have invaded more brackish conditions
and even freshwater, where modern unionid mussels
cause enormous damage as biofoulers. It is clear that
the most primitive bivalves were marine shallow
burrowers and that other life styles evolved later. It is
also apparent that many of the more specialized life
habits have evolved separately in a number of different lineages (i.e., polyphyletically). Seminal work by
S. M. Stanley firmly established how different aspects
of the morphology of living bivalves could be related
to their life habits, such that it is possible to use these
characteristics of extinct taxa to reconstruct the life
habits of fossil groups.
A large proportion of all bivalves (around 50% of all
modern families) burrow into soft sediments using the
foot. Most are equivalve and are isomyarian (i.e., the
posterior and anterior adductor muscles are of equal
size). The depth to which they burrow varies between
taxa, from those which lie just under the surface with
the edge of the shell virtually level with the sediment
water interface (e.g., Cerastoderma; Figure 6C), to
depths of several centimetres (e.g., Mya; Figure 6B),
with Panopea reaching spectacular depths of up to
1 m. The key to successful burrowing is maintaining
contact with the seawater in order to continue
both feeding and respiration. This is achieved by
the siphons, snorkel-like extensions of the posterior
mantle. The length of the siphons, and therefore the
depth of burrowing, can be inferred from the shells by
the size of the pallial sinus; deeper burrowers have
more indented pallial sinuses, whereas very shallow
burrowers have no sinus at all (see Figure 6). Very

deep burrowers, such as Mya, have siphons so long

that they are unable to withdraw them fully into the
shell when it shuts, and have a permanent posterior
siphonal gape through which they protrude. Shallow
burrowers generally have strong, robust shells, often
with a pronounced radial or concentric ornament that
may assist the burrowing process or help the animal
remain locked into the sediment. Deeper burrowers
tend to have thinner shells and are often smooth
shelled. Although the foot is never preserved, its presence may be inferred from the pedal musculature on
the inside of the valves and, in cases where the animal
is a rapid and deep burrower (such as the razor shell
Ensis), the foot may be so well developed as to require
an anterior pedal gape.
It is clear from studies of the siphons of living
bivalves that they are constructed in a number of different ways, suggesting that the deep burrowing habit
has evolved independently in several clades.
Almost all larval bivalves attach to the substrate, if
only briefly, with tanned protein threads (the byssus)
secreted by a gland at the base of the foot. In a large
number of taxa, this habit has been neotenously
retained into adulthood, and again it is clear that
this has happened repeatedly in different groups.
Byssate bivalves fall into two categories: those like
Pinna (Figure 6A) and Modiolus that are orthothetic
and live attached to clasts within the sediment in
which they are partially buried (endobyssate), and
those that are attached to the surface of hard
substrates (epibyssate), either in an orthothetic (e.g.,
Mytilus; Figure 7D) or a pleurothetic (e.g., Isognomon ephippium; Figure 7C) orientation. Orthothetic
byssate bivalves tend to be equivalve and have much
reduced anteriors. This anterior reduction is reflected374 FOSSIL INVERTEBRATES/
Figure 5 Selected hinge plates showing some of the variety of ligament insertion
and arrangement of teeth. (A) Cerastoderma edule :
heterodont dentition with two centrally placed cardinal teeth and two lateral te
eth. (B) Venus casina : heterodont (similar to Cerastoderma
but with no lateral teeth). (C) Arca tetragona : taxodont dentition with numerou
s teeth arranged in a row; the ligament forms a chevron
pattern on the broad triangular area below the umbones. (D) Chlamys varia : two
simple teeth with the internal ligament occupying a
triangular pit below the umbones.
Figure 6 Morphology and mode of life of bivalves living in or partially within s
oft substrates. (A) Pinna nobilis . Not to scale. (B) Mya
truncata . (C) Cerastoderma edula .
in the adductor musculature, which (although still
dimyarian) is heteromyarian, with the anterior adductor much smaller than the posterior (Figure 7D).
Pleurothetic byssate bivalves are often markedly
inequivalve, with the lower valve (which in the majority of cases is the right) often larger than the other.
Although they are dimyarian early in ontogeny, the
adults are monomyarian, having lost the anterior
muscle during ontogeny; the remaining posterior
muscle is often large and centrally placed (Figure 7C).
The presence of a byssus may be inferred from either a

slight gape between the valves through which it passes

(the byssal gape), or more obviously the byssal notch in
scallops (Figure 4B).FOSSIL INVERTEBRATES/Bivalves 375
Figure 7 Morphology and mode of life of bivalves living on or boring into substr
ates. (A) Spengleria rostrata . (B) Spondylus americanus .
(C) Isognomon ephippium . (D) Mytilus edulis . (E) Placuna placenta . (F) Grypha
ea arcuata . Not to scale.
Whereas byssate attachment is flexible and also
renewable, some bivalves permanently attach to
hard substrata by a calcareous cement. The cemented
habit always succeeds a byssate phase and it is clear
that it has evolved independently in a number of different clades (e.g., oysters, rudists, and a number of
pectinoids including Spondylus). Cemented bivalves
are often easily recognizable from their irregular
morphology, developed because of the requirement
to conform to substratal irregularities, and are markedly inequivalve. As they tended to evolve from
pleurothetic byssate stock, most are also monomyarian (Figure 7B). Most cement mainly by the right
valve, but a major group, the oysters, do so by the
left valve. The size of the attachment scar varies and
the substrate may be instantly recognizable; for
example, oysters were often attached to ammonite
shells, even if the scar and substrate are no longer
attached. Most cementers have thick, robust shells
and may be extravagantly ornamented with spines or
flanges (e.g., Spondylus; Figure 7B).
Free Living
A number of different taxa have independently abandoned attachment to become free living on softer
sediments. Here, the challenge is not to sink into the
substrate, and this has been solved in two ways. The
first is by adopting a snow-shoe-type morphology,
i.e., resting on a large surface area, epitomized by the
wafer-thin window pane shell Placuna (Figure 7E).
Alternatively, they may be semi-submerged in the soft
sediment like an iceberg. This strategy is inferred for
the thick-shelled devils toenail Gryphaea, common in
Mesozoic clay facies (Figure 7F). These aberrant
oysters clearly had a cemented phase, marked by a
small attachment scar at the umbo. A few free-living
bivalves (notably several groups of scallops) also
have the ability to swim short distances if they are
threatened. These have smooth hydrodynamic shells
and a well-developed posterior adductor muscle
whose vigorous contraction provides the propulsion.
A number of groups, once again polyphyletically,
have evolved to excavate burrows in hard substrates
by boring. The most successful of these, the mytilid
lithophagids, do so principally by acidic secretions
(which presents its own challenge of not dissolving
its own shell), whilst others, e.g., Pholas, bore at least
partly by physically rasping the substrate with small
projections on the outside of the shell. Some of the
most bizarre borers are the teredinids which excavate
long cylindrical boreholes in wood and have the enzymatic capability of digesting the cellulose. These
shipworms are thought to have been part of the

undoing of the ships of the Spanish Armada.

Members of the boring group as a whole have a
very varied morphology, but may be easily recognized
because they are almost invariably fossilized within
Statesman of London, called it "difficult, though not absolutely so." The Times
London, did an informal poll, asking seven reporters and a math student to read
it and
report on its accessibility. The verdict was mixed. "It all made beautiful sense
as I read it,
though it tended to vanish like a dream when I put the book down," one wrote.
Albert Einstein once said that scientific theories should be able to be describe
d so simply
that a child could understand them. Complaints that modern physics fails this st
abysmally are as old, well, as modern physics, and are not confined to the child
The story goes that when the astronomer Arthur Eddington, whose observations of
bending during a solar eclipse in 1919 confirmed Einstein s general theory of re
was congratulated by a colleague on being one of the three people in the world w
understood the abstruse theory, Eddington fell uncharacteristically silent. Chid
ed for
exhibiting a false modesty, Eddington replied, on the contrary, that he had been
trying to
imagine who the third person could be.
This newspaper s early accounts of Einstein s and Eddington s 1919 breakthroughs
focused on the theory s incomprehensibility. "Efforts made to put in words intel
ligible to
the nonscientific public the Einstein theory of light proved by the eclipse expe
dition so
far have not been very successful," began a article on Nov. 10, 1919.
Niels Bohr, one of the founders of quantum mechanics, once said that anyone who
not outraged on hearing about the theory -- with its waves acting as particles,
acting like waves, and the microscopic randomness and uncertainty it ascribed to
nature - had not really understood it.
Recent advances have made it even harder to explain the universe. The latest ver
sion of
the putative theory of everything posits a universe with 10 or 11 dimensions, in
stead of
35the 3 of space and 1 of time of everyday experience, inhabited by wriggling st
rings or
membranes. Nevertheless, scientists go on gamely trying to tell us what they are
up to, in
a book-writing tradition that includes Darwin s "Origin of Species," and Einstei
n s,
"Relativity: The Special and the General Theory," written in 1916 and never out
of print.
Part of the lure of these books is the chance to reclaim one s citizenship in a
troubled and
baffling cosmos by hearing the word from the horse s mouth, from someone who has
touched the cosmic mystery personally. But another part is surely being treated
like an
adult, of entering a rough-hewn colleagueship by being trusted to put work into

deciphering statements like the one at the beginning of this essay, or to deal w
ith straight
talk of the nature of science and the universe.
Here, for example, is Dr. Hawking about those troublesome extra dimensions requi
red by
string theory but apparently unavailable for parking cars. "I must say that pers
onally, I
have been reluctant to believe in extra dimensions," he writes on Page 54 of the
book. "But as I am a positivist, the question Do extra dimensions really exist?
has no
meaning. All one can ask is whether mathematical models with extra dimensions pr
a good description of the universe."
In other words, if the experiments come out right, it doesn t matter. This could
considered jarring if you cling to the notion that science is the search for a r
eality that is
deeper than the measurements on a laboratory table. But, quantum theory and rela
have taught us, science is about what can be observed and measured or it is abou
t nothing
at all. In science, as in democracy, there is no hidden secret knowledge, all th
at counts is
on the table, observable and falsifiable. All else is metaphysics.
When it comes to putting the goods on the table without condescending, Dr. Hawki
ng is a
genius. While many authors of science books plough through chapters full of
fundamentals before getting to the new stuff, Dr. Hawking, with perhaps a height
appreciation of time, breezes speedily to the frontier without apologies.
For those who cannot keep up, Dr. Hawking has also provided a legacy. The succes
s of
his earlier book and that of Carl Sagan s "Cosmos" are widely credited with havi
ng given
a commercial lift to the science-book genre, helping pave the way for efforts li
ke "The
Elegant Universe," by Dr. Brian Greene, a Columbia University string theorist; "
Inflationary Universe," by Dr. Alan Guth, cosmologist at the Massachusetts Insti
tute of
Technology; and "The Quark and the Jaguar," by the Nobel laureate Murray Gell-Ma
To the extent that Dr. Hawking s earlier success has spawned imitators and widen
ed the
circle of readers and their sophistication, he has engineered a kind of positive
and he has increased the odds that the readers will follow him and get to the en
d of the
book this time.
Copyright 2002 The New York Times Company
36Mysteries of the Universe
The End of Everything
In the decades that astronomers have debated the fate of the expanding universe
-whether it will all end one day in a big crunch, or whether the galaxies will sa
il apart
forever -- aficionados of eternal expansion have always been braced by its seemi

endless possibilities for development and evolution. As the Yale cosmologist Dr.
Tinsley once wrote, "I think I am tied to the idea of expanding forever."
Life and intelligence could sustain themselves indefinitely in such a universe,
even as the
stars winked out and the galaxies were all swallowed by black holes, Dr. Freeman
a physicist at the Institute for Advanced Study, argued in a landmark paper in 1
979. "If
my view of the future is correct," he wrote, "it means that the world of physics
astronomy is also inexhaustible; no matter how far we go into the future, there
always be new things happening, new information coming in, new worlds to explore
, a
constantly expanding domain of life, consciousness, and memory."
Now, however, even Dr. Dyson admits that all bets are off. If recent astronomica
Transcription Control in
Transcription in eukaryotes differs from that in
prokaryotes in two main respects. In eukaryotes, one gene codes for a single polypeptide
(monocistronic transcription unit) and the initial transcript is processed into mature messenger mRNA. This involves intron splicing (see
p. 50) and substantial modification of the ends
of the primary transcript.
A. Prototype of a eukaryotic structural
A structural gene is a gene that codes for a polypeptide gene product. It can be divided into sections involved in transcription (transcription
unit) and regulatory sequences. Regulatory
sequences are located both upstream (the 5!
direction) and downstream (the 3! direction) of
the gene. In addition, internal regulatory
sequences may occur in introns. Some regulatory sequences are located far from the gene.
Together with the promoter (see p. 206), they
are required to regulate transcription.
nosine is methylated in position 7, as are the
two initial ribose residues at the beginning of
the RNA chain. Except for the mRNAs transcribed by DNA viruses, eukaryotic mRNA usually contains a single protein-coding sequence
(monocistronic messenger).
D. Polyadenylation at the 3! end
Eukaryotic termination signals have been less
well recognized than the regulators of gene activity at the 5! end. Eukaryotic primary transcripts are split by a specific endonuclease
shortly after the sequence AAAUAA. Subsequently, about 100 250 adenine nucleotides
are attached to the 3! end of the transcript by
means of a poly(A)-polymerase (polyadenylation). The poly(A) end binds to a protein. All
mRNAs, except those that code for histone proteins, possess a poly(A) terminus.

(Figures after Singer and Berg, 1991).

Singer, M., Berg, P.: Genes & Genomes. Blackwell
Scientific, Oxford, 1991.
B. Prototype of mature eukaryotic
Mature eukaryotic mRNA is produced from its
precursor RNA by the removal of introns, addition of a 5! cap at the 5! end, and addition of
numerous adenine nucleotides at the 3! end
(polyadenylation). A noncoding sequence (5!
leader) is located in front of the translation start
signal (AUG), and a trailer sequence, at the 3!
end in back of the translation stop signal (UAA).
Both addition of the 5! cap and polyadenylation
involve enzymatic reactions.
C. 7-Methyl-guanosine cap
The translation of eukaryotic mRNA is similar to
that of prokaryotic mRNA, with two distinct
differences: (1) transcription and translation
occur at different locations in the eukaryotic
cell: transcription occurs in the cell nucleus,
and translation in the cytoplasm; (2) the 5! and
3! ends of eukaryotic mRNA have special structures. The structure at the 5! end is called a cap.
Through the action of guanosine-7-methyltransferase, guanosine is bound by a triphosphate bridge to the first and second ribose
groups of the precursor mRNA chain. The guaPassarge, Color Atlas of Genetics 2001 Thieme
All rights reserved. Usage subject to terms and conditions of license.Transcript
ion Control in Eukaryotes
Passarge, Color Atlas of Genetics 2001 Thieme
All rights reserved. Usage subject to terms and conditions of license.
Regulation of Gene Expression
in Eukaryotes
Precisely regulated gene expression is a prerequisite for producing and maintaining the
many different types of cells and tissues of a
multicellular organism. Cells differentiate into
their particular cell types by means of combinations of expressed and repressed genes. During
differentiation the tightly regulated genes function in the order, usually sequential, required
for a particular cell fate (developmental pathways). Many regulator genes and their proteins
have been identified (cf. part III, Genetics and
Medicine). The following outlines some important principles of the specific control of gene expression in eukaryotic cells.
A. Levels of control of eukaryotic gene
In principle, expression can be regulated at four
distinct levels. The first and by far the most important is primary control of transcription.
Processing to mature RNA can be regulated at
the level of the primary RNA transcript. A
frequently observed process is alternative splic-

ing (see D). Translation can be varied by RNA

editing (see B for an example). At the protein
level, posttranslational modifications can determine the activity of a protein. The cleavage of
preproinsulin to form mature insulin, glycosylation or hydroxylation, and protein folding are
some examples (see p. 32, 362).
B. RNA editing
RNA editing modifies genetic information at the
RNA level. An important example is the apolipoprotein-B gene involved in lipid metabolism. It
encodes a protein of 4538 amino acids, apolipoprotein B. This is synthesized in the liver and
secreted into the blood, where it transports
lipids. A related shorter form of the protein with
2153 amino acids, Apo B-48 (250 kDa, instead of
512 kDa for Apo B-100), is synthesized in the intestine. An intestinal deaminase converts a cytosine in codon 2158, CAA (glutamine), to uracil
(UAA). This change results in a stop codon (UAA)
and thereby terminates translation at this site.
C. Long-range gene activation by an
Enhancers control gene activity at a distance.
An enhancer is a distant site involved in initiation of transcription (see p. 206). It may be located either upstream or downstream of the
same DNA strand (cis-acting) or on a different
DNA strand (trans-acting). Enhancer elements
provide tissue-specific or time-dependent regulation. It is unclear how enhancers can exert
their effect from a considerable distance. One
model suggests that DNA forms a loop between
enhancer and promoter. Activator proteins
bound to the enhancer, e.g., a steroid hormone,
could then come into contact with the general
transcription factor complex at the promoter.
Others might function as repressors (cf. transcription control in prokaryotes, p. 210).
D. Alternative RNA Splicing
A DNA segment can code for different forms of
mRNA when different introns are removed from
the primary transcript (alternative splicing). By
means of alternative gene splicing, a gene can
code for different, albeit similar gene products.
This allows a high degree of functional flexibility. Numerous examples of differential RNA
splicing are known for mammalian genes. For
example, the primary transcript for the calcitonin gene contains six exons. They are spliced
into two different types of mature mRNA. One,
consisting of exons 1 4 (but not exons 5 and 6),
is produced in the thyroid and codes for calcitonin. The other consists of exons 1, 2, 3, 5, and
6, but not exon 4. It codes for a calcitonin-like
protein in the hypothalamus (calcitonin generelated product, CGRP).
Alberts, B., et al: Essential Cell Biology. An Introduction to the Molecular Biology of the Cell.

Garland Publishing Co., New York, 1998.

Lewin, B.: Genes VII, Oxford Univ. Press, Oxford,
Blackwood E.M., Kadonga J.F.: Going the distance: A current view of enhancer action.
Science 281 :60 63, 1998.
Stryer, L.: Biochemistry. 4 th ed. W.H. Freeman &
Co, New York, 1995.
Watson J.D., et al.: Molecular Biology of the
Gene. 4 th ed. Benjamin/Cummings Publishing Co., Menlo Park, California, 1987.
Passarge, Color Atlas of Genetics 2001 Thieme
All rights reserved. Usage subject to terms and conditions of license.217
Regulation of Gene Expression in Eukaryotes
Apo B-100
mRNA unedited
Cytosine deamination
by intestinal deaminase
Apo B-84
B. RNA editing
Activator protein
Control of
start site
Binding of an
activator protein
to the transcription
Control of
(mRNA editing)
Transcription factors
Control of
protein activity

Control of
Primary transcript
A. Levels of control of eukaryotic
gene expression
Promoter with
factors and
RNA polymerase II
C. Long-range gene activation by an enhancer
Calcitonin gene
Exon 1
Exon 2
Exon 3
Primary RNA transcript
C cells in thyroid
Exon 4 Exon 5
4 5
Exon 6 3
6 3
RNA processing
gene products
D. Alternative RNA splicing
Passarge, Color Atlas of Genetics 2001 Thieme
All rights reserved. Usage subject to terms and conditions of license.
3 218

DNA-Binding Proteins
Regulatory DNA sequences interact with proteins to exert proper functional control. Regulatory proteins can recognize specific DNA
sequences because the surface of the proteins
fits precisely onto the DNA surface. Three basic
groups of regulatory DNA sequences can be distinguished: (1) sequences that establish the
exact beginning of translation; (2) DNA segments that regulate the end, or termination;
and (3) DNA sequences near the promoter that
have specific effects on gene activity (repressors, activators, enhancers, and others).
A. Binding of a regulatory protein to
Gene regulatory proteins can recognize DNA
sequence information without having to open
the hydrogen bonds within the helix. Each base
pair represents a distinctive pattern of hydrogen bond donors (example shown in red) and
hydrogen acceptors (example shown in green).
These proteins recognize the major groove of
DNA, where binding takes place. Here a single
contact of an asparagine (Asn) of a gene-regulatory protein with a DNA base adenine (A) is
shown. A typical area of surface-to-surface contact involves 10 20 such interactions. (Figure
redrawn from Alberts et al., 1998, p. 276).
B. An helix inserts into  mjor
groove of opertor DNA
One prt of the protein, n helix (the
sequence-reding or recognition helix) is inserted into the mjor groove of DNA. Here the
sequence Q-Q-Q-S-T (glutmine Q, serine S,
threonine T) in the recognition sequence of the
bcteriophge 434 repressor bonds with
specific bses in  mjor groove of opertor
DNA. (Figure redrwn from Lodish et l., 2000,
p. 351).
C. Zinc finger motif
Another group of proteins re clled zinc fingers
becuse they resemble fingers (see D). They re
involved in importnt functions during embryonic development nd differentition. The bsic
zinc finger motif consists of  zinc tom connected to four mino cids of  polypeptide
chin. Here, two histidine (H) nd two cysteine
(C) residues re shown in the schem on the
left. The three-dimensionl structure on the
right consists of n ntiprllel sheet (amino
acids 1 10), an helix (mino cids 12 24),
nd the zinc connection. Four mino cids, cysteines 3 nd 6 nd histidines 19 nd 23, re
bonded to the zinc tom nd hold the crboxy
(COOH) end of the helix to one end of the
sheet. (Figure redrawn from Alerts et al., 1994,
p. 411).
D. Zinc finger proteins ind to DNA
The interaction with DNA is strong and specific.
Each protein recognizes a specific DNA

sequence. As the numer of zinc fingers can e

varied, this type of DNA-inding has great evolutionary flexiility. (Figure redrawn from Alerts et al. 1994).
E. Binding to a response element
Many hormones and growth factors activate
cell-surface receptors. In contrast, steroid hormones enter the target cells and interact there
with a specific receptor protein in the cytosol.
The hormonereceptor complex then migrates
to specific sites of DNA. The hormone-inding
domain will prevent inding to DNA unless the
hormone is present. Activated receptors ind to
specific DNA sequences called hormone response elements (HREs). Each polypeptide
chain of the receptor contains a zinc atom
ound to four cysteines (1). The skeletal model
shows the two DNA-inding domains inding
to the HRE in two adjacent major grooves of the
target DNA (2). The space-filling model shows
how tightly the recognition helix of each dimer
of this protein fits into the major groove of DNA
shown in red and green (3). (Figure redrawn
from Stryer, 1995, p. 1002).
Alerts, B., et al.: Molecular Biology of the Cell.
3 rd ed. Garland Pulishing Co., New York,
Alerts, B., et al.: Essential Cell Biology. Garland
Pulishing Co., New York, 1998.
Lodish, H., et al.: Molecular Cell Biology. 4 th ed.
Scientific American Books, F.H. Freeman &
Co., New York, 2000.
Stryer, L.: Biochemistry. 4 th ed. W.H. Freeman &
Co., New York, 1995.
Tjian, R.: Molecular machines that control
genes. Sci. Am. 272 :38 45, 1995.
Passarge, Color Atlas of Genetics 2001 Thieme
All rights reserved. Usage suject to terms and conditions of license.DNA-Bindin
g Proteins
DNA-inding protein
Major groove
CH 3
CH 2
To sugar

To sugar
Minor groove
A. Binding of a regulatory protein to DNA
H Acceptor
B. An ! helix inserts into a major
groove of operator DNA
NH 2
6 C
H 19
His 19
F 10
D. A zinc finger protein inds to DNA

H 2 N
C. Zinc finger motif
E. Binding to a response element
Passarge, Color Atlas of Genetics 2001 Thieme
All rights reserved. Usage suject to terms and conditions of license.
Other Transcription Activators
Transcription activators are dimeric proteins
with distinct functional domains: a DNA-inding domain and an activation domain. The DNAinding domain interacts with specific regulatory DNA sequences. The activation domain interacts with other proteins that stimulate transcription. Transcription activators participate in
the assemly of the initiation complex, for example, y stimulating the inding of transcription factor IID (TFIID, see p. 212) to the promoter. Other activators may interact with
general transcription factors. They provide a
second level of transcriptional control.
A. Leucine zipper dimer
Most DNA-inding regulatory proteins recognize specific sites as dimers. One part of the
molecule serves as the recognition molecule,
the other stailizes the structure. A particularly
striking example is given y proteins with a
leucine zipper motif. The name is derived from
the asic structure. Two helices re joined like
 zipper by periodiclly repeted leucine residues locted t the interfce of the two helices.
The two helices seprte, form  Y-shped
structure, nd extend into the mjor groove of
the DNA (1). Leucine zipper proteins my be homodimers with identicl subunits (2, 3) or heterodimers with different lbeit similr subunits
(4). The bility to form unlike dimers (heterodimeriztion) gretly expnds the spectrum of
specificites. The use of combintions of different proteins to control cellulr functions is
clled combintoril control. (Figure redrwn
from Alberts et l., 1994).
A DNA-binding motif relted to the leucine zipper is the helixloophelix (HLH) motif (not
shown). The HLH motif consists of one short

helix nd one longer helix. The two helices

re connected by  flexible loop of protein.
ment re regulted by steroids (steroid-responsive trnscription). The ltter include glucocorticoids nd minerlocorticoids, the steroids of
glycogen nd minerl metbolism; sex hormones, which function in embryonic sex differentition nd control of reproduction; nd
others. Norml bone development nd function
re under the control of steroidlike vitmin D.
Another steroidlike hormone is retinoic cid, n
importnt regultor of differentition during
embryogenesis (morphogen). These hormones
initite their physiologicl effects by ssocition with corresponding steroid-specific trnscellulr receptors (hormonereceptor complex).
C. Evidence of  protein-binding
region in DNA
Protein-binding regions in DNA represent regultory res; thus, their nlysis cn yield some
insights into gene regultion. Protein-binding
DNA regions cn be demonstrted in severl
wys. With bnd-shift nlysis (1), proteinbound nd non-protein-bound DNA frgments
re differentited using gel electrophoresis in
direction towrds the smll frgments,  DNA
frgment tht is prt of  DNA-protein complex
migrtes more slowly thn  free DNA frgment
of the sme size. The DNA-protein complex is
found t  different position (bnd shift). DNA
footprinting (2) is nother procedure for identifying protein-binding sites on DNA. The principle of DNA footprinting is tht  proteinbound DNA region, e.g., the polymerse-promoter complex, is protected from the effects of
 DNA-cleving enzyme (DNAse I). Previously
isolted DNA is cut into different frgments by
DNAse, nd the frgments re sorted ccording to size by gel electrophoresis. Since the DNA
protein-binding region is protected from clevge by DNAse I (DNAse I protection experiment), DNA bnds from the binding region re
missing (footprint).
B. Activtion by steroid hormone
binding References
Trnscriptionl enhncers re regultory regions of DNA tht increse the rte of trnscription. Their spcing nd orienttion vry reltive
to the strting point of trnscription. An enhncer is ctivted by binding to  hormonereceptor complex. This ctivtes the promoter,
nd trnscription begins (ctive gene). Numerous importnt genes in mmmlin develop- Alberts, B., et l.: Moleculr Biology
of the Cell.
3 rd ed. Grlnd Publishing Co., New York,
Alberts, B., et l.: Essentil Cell Biology. Grlnd
Publishing, Co., New York, 1998.
Lodish, H., et l.: Moleculr Cell Biology. 4 th ed.

Scientific Americn Books, F.H. Freemn &

Co., New York, 1999.
Pssrge, Color Atls of Genetics 2001 Thieme
All rights reserved. Usge subject to terms nd conditions of license.Other Trn
scription Activtors
Pssrge, Color Atls of Genetics 2001 Thieme
All rights reserved. Usge subject to terms nd conditions of license.
Inhibitors of Trnscription nd
A number of nturl nd rtificil substnces
re ble to inhibit trnscription or trnsltion.
They cn be used to tret cncer or s ntibiotics
to tret infections. Although most substnces
re unspecific nd not suitble for therpeutic
purposes, some re very specific nd therefore
importnt for the understnding of trnscription nd trnsltion or for therpy. Bsiclly,
one cn distinguish whether n gent interferes
with trnscription or with trnsltion.
A. Insertion of ctinomycin D
between  GC bse pir
Actinomycin D is  complex polypeptide produced by  species of streptomyces bcteri. It
consists of  phenoxzone ring with two symmetricl side chins (1). It cts by becoming interclted between two neighboring GC bse
pirs in double-strnded DNA. Viewed from the
side (2), the inserted ctinomycin D molecule is
seen very distinctly within the DNA double
helix. In the view from bove (3) the ctinomycin D molecule forms  nrrow lyer within the
DNA double helix, bound by the two neighboring GC bse pirs. The degree of inhibition by
ctinomycin D vries gretly. High concentrtions of ctinomycin D block repliction,
wheres low concentrtions suffice to inhibit
bosome, nd the protein synthesis ends premturely.
(Figures A nd B from Singer nd Berg, 1991).
C. Inhibitors of protein synthesis
Numerous nturlly occurring nd rtificilly
produced substnces inhibit protein synthesis
by inhibiting trnscription or certin phses of
trnsltion. Some hve clinicl significnce s
ntibiotics; others re toxicologiclly significnt. An exmple for the specificity of some inhibitors is -mnitin,  dicyclic octpeptide of
the fungus Amnit phlloides. In very low concentrtions, it binds to RNA polymerse II nd
thereby blocks the formtion of precursor
mRNA in eukryotes. In contrst, RNA polymerse I is insensitive to this toxin, nd polymerse III binds to it only in higher concentrtions. (Dt fter Singer nd Berg, 1991).
Singer, M., Berg, P.: Genes & Genomes. Blckwell
Scientific, Oxford, 1991.

B. Puromycin imittes n minocyl

Puromycin,  polypeptide from Streptomyces lboniger, blocks polypeptide synthesis in the ribosomes of prokryotes nd eukryotes. Its ction is bsed on the structurl similrity with n
minocyl tRNA. An minocyl tRNA is  tRNA
molecule with n mino cid ttched to its 3!
end. Normlly  peptide bond is formed by peptidyltrnsferse between the mino group of
the incoming minocyl tRNA t the A (minocyl) position nd the crboxyl group of the
peptidyl tRNA t the P (peptidyl) position. The
structure of puromycin resembles tht of
minocyl tRNA, but lcking n interction with
the codon it cnnot be ttched to the A position in the ribosome. The resulting polypeptidylpuromycin dduct is removed from the riPssrge, Color Atls of Genetics 2001 Thieme
All rights reserved. Usge subject to terms nd conditions of license.Inhibitors
of Trnscription nd Trnsltion
Pssrge, Color Atls of Genetics 2001 Thieme
All rights reserved. Usge subject to terms nd conditions of license.
DNA Methyltion
Methyltion of cytosine residues in DNA plys
n importnt role in gene regultion. DNA
methyltion is required for norml embryonic
development. Genomic imprinting, X chromosome inctivtion, chromtin modifiction, nd
silencing of endogenous retroviruses ll depend
on estblishing nd mintining proper methyltion ptterns. DNA methyltion is genespecific nd occurs genome-wide. Two types of
methyltrnsferse cn be distinguished by their
bsic functions: mintennce methyltion nd
de novo methyltion.
A. Mintennce methyltion
This type of methyltion is responsible for
dding methyl groups to the newly synthesized
DNA strnd fter repliction nd cell division.
The methylted sites in the prentl DNA (1)
fter repliction (2) serve s templte for correct methyltion of the two new strnds. This
ensures tht the previous methyltion pttern
is correctly mintined (3). It results in both
dughter strnds being methylted t the sme
sites s the prentl DNA. The enzyme responsible for this is Dnmt1 (DNA methylse 1).
Its function is essentil. Mice deficient in this
enzyme die s  result of genome-wide demethyltion.
B. Recognition of  methylted DNA
Certin restriction enzymes do not cleve DNA
when their recognition sequence is methylted
(1). The enzyme HpII cleves DNA only when
its recognition sequence 5!-CCGG-3! is not
methylted (2). MspI recognizes the sme 5!-

CCGG-3! sequence irrespective of methyltion

nd cleves DNA t this site every time. This
difference in clevge pttern results in DNA
frgments of different sizes serves to distinguish the methyltion pttern of the DNA.
C. DNA methyltion de novo
This is the second type of DNA methyltion.
Here methyl groups re dded t new positions
of both strnds of DNA, not just in the hemimethylted strnd s in mintennce methyltion shown in A. Two genes for different
methyltrnsferses with overlpping functions
in globl remethyltion hve been identified recently: Dnmt3 nd Dnmt3b. Unmethylted
DNA (1) is methylted by their enzymes (2) in 
site-specific nd tissue-specific mnner (3).
Trgeted homozygous disruption of the mouse
Dnmt3 nd Dnmt3b genes results in severe
developmentl defects. Double homozygous
mutnts die before dy 11.5 of the 21-dy
embryonic development.
D. Humn DNMT3B gene
Muttions in the humn gene DNMT3B encoding type 3B de novo methyltrnsferse cuse 
distinctive disese clled ICF syndrome (immunodeficiency, centromeric chromosoml instbility, nd fcil nomlies, McKusick ctlog
no. 242860). The centromeres of chromosomes
1, 9 nd 16, where stellite DNA types 2 nd 3
re locted, re unstble. Clssicl stellite DNA
is grossly undermethylted in ll tissues. The
humn gene (1) consists of 23 exons spnning
47 kb of genomic DNA. Six exons re subject to
lterntive splicing. The protein (2) consists of
845 mino cids with five DNA methyltrnsferse motifs (I, IV, V, IX, X) in the C-terminl region. The rrows point to six different muttions. The muttion t position 809 (3), 
chnge of A to G in codon 809, i.e., GAC (Asp) to
GGC (Gly), leds to the replcement of sprgine (Asn) by glycine (Gly). Both prents
re heterozygous for this muttion.
(Figure dpted form Xu et l., 1999).
Bird, A: DNA methyltion de novo. Science
286 :2287 2288, 1999.
Hnsen, R. S. , et l.: DNMT3B DNA methyltrnsferse gene is mutted in the ICF immunodeficiency syndrome. Proc. Nt. Acd.
Sci. 96 :14 412 14 417, 1999.
Okno, M., et l.: DNA Methyltrnsferses
Dnmt3 nd Dnmt3b re essentil for de
novo methyltion nd mmmlin development. Cell 99 :247 257, 1999.
Reik, W., Kelsey, G., Wlter, J.: Dissecting de novo
methyltion. Nture Genet. 23 : 380 382,
Robertson, K.D., Wolffe, A.P.: DNA methyltion
in helth nd disese. Nture Reviews 1 :11
19, 2000.

Xu, G., Bestor, T., et l.: Chromosome instbility

nd immunodeficiency syndrome cused
by muttions in  DNA methyltrnsferse
gene. Nture 402 :187 191, 1999.
Pssrge, Color Atls of Genetics 2001 Thieme
All rights reserved. Usge subject to terms nd conditions of license.DNA Methyl
Pssrge, Color Atls of Genetics 2001 Thieme
All rights reserved. Usge subject to terms nd conditions of license.
Genomic Imprinting
Genetic contributions from both the mternl
nd the pternl sets of chromosomes re necessry for  mmmlin zygote to develop normlly. The reson lies in  prent-of-originspecific expression of certin genes. The
genome contins defined regions where only
the mternl gene copy is expressed, nd not
the pternl copy, nd vice vers. This llelespecific gene expression, depending on the
prentl origin, results from so-clled genomic
imprinting. Genomic imprinting hs importnt
implictions for humn genetic disese (see
p. 398).
A. The importnce of two different
prentl genomes
Different developmentl results re observed
when the femle pronucleus is removed from 
mouse zygote (1) before the pronuclei hve
fused nd is replced either by nother mle
pronucleus (2) or by  control (i.e., the sme s
removed) (3), or when the mle pronucleus is
removed nd replced by  femle pronucleus
(4) or  control. If the femle pronucleus is replced by  mle pronucleus (2), the zygote first
ppers norml. However, fter implnttion
should ensue, nerly ll ndrogenotes fil to
complete preimplnttion (2). Those few tht,
completely bnormlly. Such embryos do not
develop beyond the 12-somite stge.
When  mle pronucleus is replced by  femle
pronucleus (4),  gynogenetic zygote, which
differs mrkedly from the ndrogenote. Here,
bout 85% of gynogenotes rech norml preimplnttion development. But lthough the embryo t first develops firly well, the extrembryonic membrnes re bsent or underdeveloped, nd the gynogenote dies before
reching the 40-somite stge. (Figure redrwn
from Spienz nd Hll, 1995).
B. Humn embryonic development
depends on the presence of 
mternl nd  pternl genome
A nturlly occurring humn ndrogenetic zygote is  hydtidiform mole (1). This is n bnor-

ml plcentl formtion contining two sets of

pternl chromosomes nd none from the
mother. No embryo develops lthough implnttion tkes plce. The plcentl tissues develop
mny cysts (2). When only mternl chromosomes re present, n ovrin tertom with
mny different types of fetl tissues develops
(3). No plcentl tissue is present in this nturlly occurring gynogenetic zygote. In triploidy,  reltively frequent globl chromosoml lethl humn disorder (see p. 402), extreme hypoplsi of the plcent nd fetus is
observed when the dditionl chromosoml set
is of mternl origin (4). (Photogrphs kindly
provided by Professor Helg Rehder, Mrburg.)
C. Genomic imprinting is estblished in
erly embryonic development
The imprint pttern present in somtic cells (1),
with one llele ctive only either, the mternl
or the pternl, propgted through ll mitotic
divisions. However, in primordil germ cells the
imprint is ersed (2). During gmetogenesis the
imprint is reset (3). In the mle germline ll
gmetes receive the pternl imprint nd in the
femle germline ll gmetes receive the mternl imprint. After fertiliztion, the correct imprint pttern is present in the zygote (4). It is
mintined through ll subsequent cell divisions under the control of  regionl imprinting
center (see p. 398).
Brlow, D.P.: Gmetic imprinting in mmmls,
Science 270 :1610 1613, 1995.
Horsthemke, B.: Genomisches Imprinting beim
Menschen: Grundlgen und klinische Relevnz. Biospektrum 4 :23 26, 1998.
Morrison, I.M., Reeve, A.E.: Ctlogue of imprinted genes nd prent-of-origin effects
in humns nd nimls. Hum. Mol. Genet.
7 :1599 1609, 1998.
Reik, W., Surni, A., eds.: Genomic Imprinting.
IRL Press t Oxford University Press, Oxford,
Spienz, C., Hll, J.G.: Genetic imprinting in
humn disese. pp. 437 458. In: The Metbolic nd Moleculr Bses of Inherited Disese. 7 th ed. C.R. Scriver, et l., eds. McGrwHill, New York, 1995.
Surni, A.: Imprinting nd the initition of gene
silencing. Cell 93 :309 312, 1998.
Tilghmn, S. M.: The sins of the fthers nd
mothers: Genomic imprinting in mmmlin development. Cell 96 :185 193,
Pssrge, Color Atls of Genetics 2001 Thieme
All rights reserved. Usge subject to terms nd conditions of license.227
Genomic Imprinting
dies very

Diploid zygote
Fetus bsent
or stunted
filure in most
Fetus norml
Fetus norml
until 40 somite
Preimplnttion norml,
tissues underdeveloped
A. The importnce of two different prentl genomes
Two pternl genomes
1. Hydtidiform mole
Two mternl genomes
2. Hydtidiform mole
3. Ovrin tertom
4. Triploidy 69, XXX
B. Humn embryonic development depends on presence of  mternl nd  pternl
Somtic cells
XX nd XY
P Pternl
Inctive Active
Active Inctive
M Mternl
germ cells

C. Genomic imprinting is estblished in erly embryonic development
Pssrge, Color Atls of Genetics 2001 Thieme
All rights reserved. Usge subject to terms nd conditions of license.
X-Chromosome Inctivtion
During erly embryonic development of mmmlin femles, one of the two X chromosomes
becomes inctivted. The inctivtion is induced by  gene (XIST, X-inctivtion-specific
trnscript) on the proximl long rm of the X
chromosome, trnscribed from the inctive X.
X inctivtion is  mechnism to blnce Xchromosoml gene expression between femle
nd mle cells.
A. X chromtin
In 1949, Brr und Bertrm observed  stinble
ppendge in the nucleus of nerve cells of
femle (1 nd 3) but not of mle cts (2). The
uthors nmed this structure s sex chromtin. Similr structures were found drumsticks in peripherl blood leukocytes (4) nd
smll peripherl bodies in the nuclei of fibroblsts nd orl mucos cells (5) in humns. Ech
of these structures represents one of the two X
chromosomes nd re referred to s X chromtin. (Figures 1 3 from Brr nd Bertrm,
B. Scheme of X inctivtion
Rndom inctivtion of most of the genes of one
of the two X chromosomes in femle cells
occurs erly in embryogenesis, t bout dy 21
in humns. In  given cell, it involves the X chromosome of either mternl or pternl origin.
The inctivtion pttern is normlly irreversible
nd stbly trnsmitted to ll dughter cells. The
expected distribution is usully 1 : 1. In rre instnces this my be skewed towrd  preferentil type of inctivtion. In extreme cses this
my result in clinicl mnifesttion in  heterozygous femle if the mjority of cells contin
the muttion on the ctive X chromosome.
C. Mosic pttern of expression
Femle somtic tissues show  mosiclike distribution of cells expressing just one of the two
lleles. In mice, X-chromosoml cot-color mutnts show  mosic of light- nd drk-colored
cot ptches (1, fter Thompson, 1965). In
humns,  similr distribution of norml nd
bsent swet pores is seen in femle heterozygotes for hypohidrotic ectoderml dysplsi.
The swet pores of hemizygotes re bsent
(hypohidrosis). Fingerprints of femle hetero-

zygotes show res with norml swet pores

(blck points) nd res with bsent swet pores
(2 nd 2b, from Pssrge nd Fries, 1973). In
cell cultures from femle heterozygotes for
X-chromosoml HGPRT deficiency (hypoxnthineguninephosphoribosyltrnsferse)
the colonies re either HGPRT or HGPRT + (3).
(From Migeon, 1971, with kind permission of the
uthor nd publisher).
D. Exceptions from X-inctivtion
Trnscriptionl silencing of X-chromosoml
genes in mmmlin femle cells is not
complete. Some genes escpe inctivtion nd
re expressed from both the ctive nd the inctive X chromosome. The mjority of these genes
hve  homologue on the Y chromosome,
reflecting  common evolutionry origin. Pnel
D shows genes tht re expressed on the inctive humn X chromosome. Most re locted t
the ends of the X chromosomes. (Figure
dpted from Brown et l., 1997).
E. X-inctivtion profile
An nlysis of 224 X-linked trnscripts showed
tht 34 escpe inctivtion (Crrel et l., 1999).
Of these, 31 mp to the short rm of the X chromosome. The expressed genes re open circles,
the inctivted genes filled circles. Severl
genes in the pseudoutosoml regi
observtions re correct, the future of life nd the universe will be fr bleke
In the lst four yers stronomers hve reported evidence tht the expnsion of
universe is not just continuing but is speeding up, under the influence of  mys
"drk energy," n ntigrvity tht seems to be embedded in spce itself. If tht
is true nd
the universe goes on ccelerting, stronomers sy, rther thn costing gently
into the
night, distnt glxies will eventully be moving prt so quickly tht they cn
communicte with one nother. In effect, it would be like living in the middle o
f  blck
hole tht kept getting emptier nd colder.
In such  universe, some physicists sy, the usul methods of formulting physic
s my
not ll pply. Insted of new worlds coming into view, old ones would constntly
disppering over the horizon, lost from view forever.
Cosmologicl knowledge would be frgmented, with different observers doomed to
seeing different pieces of the puzzle nd no single observer ble to know the f
te of the
whole universe or rrive t  theory of physics tht ws more thn pproximte.
37"There would be  lot of things bout the universe tht we simply couldn't pre
dict," sid
Dr. Thoms Bnks,  physicist t the University of Cliforni t Snt Cruz.
And perhps most importnt, strved finlly of the energy even to complete  tho
ught or
 computtion, the domin of life nd intelligence would not expnd, but constri
ct nd
eventully vnish like  dwindling echo into the silence of eternity. "I find th

e fte of 
universe tht is ccelerting forever not very ppeling," sid Dr. Edwrd Witte
theorist t the Institute for Advnced Study.
Tht is n understtement, in the view of Dr. Lwrence M. Kruss, n strophysic
ist t
Cse Western Reserve University in Clevelnd, who long with his collegue Dr. G
D. Strkmn hs recently tried to limn the possibilities of the fr future. An 
universe "would be the worst possible universe, both for the qulity nd quntit
y of life,"
Dr. Kruss sid, dding: "All our knowledge, civiliztion nd culture re destin
ed to be
forgotten. There's no long-term future."
Einstein's Lst Lugh
Until bout four yers go, n overwhelming prepondernce of stronomers subscri
bed to
the view tht the cosmic expnsion ws probbly slowing down becuse of the coll
grvity of the glxies nd everything else in the universe, the wy  hndful o
f stones
tossed in the ir grdully slow their scent. The only question ws whether the
hd enough grvittionl oomph to stop expnding nd bring itself bck together
in  "big
crunch," or whether the glxies would sil ever more slowly outwrd forever.
It ws to mesure tht rte of slowing of this outwrd flight, nd thus find the
nd elusive nswer to the cosmic question, tht two tems of stronomers strted
competing projects in the 1990's using distnt exploding strs, supernovs, s c
In 1998 the two tems nnounced tht insted of the expected slowing, the glxi
ctully seem to hve speeded up over the lst five or six billion yers, s if
some "drk
energy" ws pushing them outwrd.
"It's definitely the strngest experimentl finding since I've been in physics,"
Dr. Witten
sid. "People find it difficult to ccept. I've stopped expecting tht the findi
ng will be
proved wrong, but it's n extremely uncomfortble result."
To stronomers this drk energy bers  hunting resemblnce to n ide tht Alb
Einstein hd bck in 1917 nd then bndoned, lter clling it his biggest blund
er. In tht
yer he inserted  mthemticl fudge fctor tht cme to be known s the cosmol
constnt into his equtions of generl reltivity in order to stbilize the univ
erse ginst
collpse; Einstein's constnt cted s  kind of cosmic repulsion to blnce the
grvittionl pull of the glxies on one nother.
38Einstein gve up the cosmologicl constnt fter the Americn stronomer Edwin
discovered tht the universe ws expnding nd thus did not need stbilizing. Bu
t his
fudge fctor refused to die. It gined  new identity with the dvent of quntum
mechnics, the bizrre-sounding rules tht govern the subtomic relm. According

those rules, empty spce is not empty, but rther foming with energy. Inserted
Einstein's equtions, this energy would ct like  cosmologicl constnt, nd tr
y to blow
the universe prt.
According to stronomers the recently discovered drk energy now ccounts for b
two-thirds of the mss of the universe. But is this Einstein's old fudge fctor,
cosmologicl constnt, come home to roost -- in which cse the universe will cc
eternlly? Or is the presumed ccelertion only temporry, driven by one of the
mysterious force fields, dubbed quintessence, llowed by vrious theories of hig
h energy
Or is the ccelertion even rel?
"It's importnt to find out if the cosmologicl constnt is relly constnt," s
id Dr. Witten.
Becuse the repulsive force resides in spce itself, s the universe grows, the
push from
drk energy grows s well. "If drk energy is the cosmologicl constnt then it
property of the vcuum tht will lwys be with us, getting more powerful s the
gets bigger nd the universe will expnd forever," explined Dr. Adm Riess of t
he Spce
Telescope Science Institute in Bltimore. But if the drk energy is some form of
quintessence, "then there my be more such fields which rise in the future, pos
sibly of
the opposite sign, nd then ll bets re off for the future of the universe."
Dr. Kruss sid, "The good news is tht we cn't prove tht this is the worst of
possible universes."
The Long Goodbye
It might seem strnge or presumptuous for stronomers to try to describe events
ll the
wy to the end of time when physicists re still groping for  "theory of everyt
hing." But
to Dr. Kruss, this is testimony to the power of ordinry physics. "We cn still
ultimte limits on things without even knowing the ultimte theory," he sid. "W
e cn put
limits on things bsed on ordinry physics."
Dr. Dyson sid his venture into eschtology ws inspired prtly by  1977 pper
on the
future of n ever expnding universe by Dr. J. N. Islm, now t the University o
Chittgong in Bngldesh, in The Qurterly Journl of the Royl Astronomicl Soc
Dr. Dyson ws lso motivted, he wrote in his pper, to provide  counterpoint t
fmously dour sttement by Dr. Steven Weinberg, who wrote in his book "The First
Three Minutes," "The more the universe seems comprehensible, the more it lso se
39Dr. Dyson wrote, "If Weinberg is speking for the 20th century, I prefer the 1
If the present trend of ccelertion continues this is the forecst:
In bout two billion yers Erth will become uninhbitble s  grdully wrmin

g Sun
produces  runwy greenhouse effect. In five billion yers the Sun will swell u
p nd die,
burning the Erth to  crisp in the process. At bout the sme time the Milky W
y will
collide with its twin the Andromed glxy, now bout two million light-yers w
y nd
closing fst, spewing strs, gs nd plnets cross interglctic spce.
Any civiliztion tht mnged to survive these events would fce  future of inc
ignornce nd drkness s the ccelerting cosmic expnsion rushes most of the u
wy from us. "Our bility to know bout the universe will decrese with time,"
sid Dr.
Kruss. "The longer you wit, the less you see, the opposite of wht we lwys t
As he explins it, the disppernce of the universe is  grdul process. The f
glxy flies wy from us, the dimmer nd dimmer it will pper, s its light is
"redshifted" to lower frequencies nd energies, the wy  police siren sounds lo
wer when
it is receding. When it reches the speed of light, the glxy will pper to "f
reeze," like 
dncer cught in midir in  photogrph, in ccordnce to Einstein's theory of r
nd we will never see it get older, sid Dr. Abrhm Loeb, n stronomer t Hrv
Rther it will simply seem dimmer. The frther wy n object is in the sky, he
sid, the
younger it will pper s it fdes out of sight. "There is  finite mount of in
formtion we
cn collect from the universe," Dr. Loeb sid. About 150 billion yers from now
ll of the glxies in the universe will be receding fst enough to be invisible
from the
Milky Wy. The exceptions will be glxies tht re grvittionlly bound to the
cloud of
glxies, known s the Locl Group, to which the Milky Wy belongs. Within this
life would look much the sme t first. There would be glxies in the sky. "Whe
n you
look t the night the strs will still be there," sid Dr. Kruss. "To the stro
nomer who
wnts to see beyond, the sky will be sdly empty. Lovers won't be disturbed -- s
will be."
But bout 100 trillion yers from now, when the interstellr gs nd dust from w
hich new
strs condense is finlly used up, new strs will cese to be born. From tht ti
me on, the
sky will grow drker nd drker. The glxies themselves, stronomers sy, will
in blck holes within bout 1030 yers.
But even  blck hole is not forever, s Dr. Stephen Hwking, the Cmbridge Univ
physicist nd best-selling uthor, showed in pth-breking clcultions bck in
Applying the principles of quntum mechnics to these dred-sounding objects, Dr
Hwking discovered tht  blck hole's surfce, its so-clled event horizon, wou

fluctute nd exude energy in the form of rndom bursts of prticles nd rditi
growing hotter nd hotter until the blck hole eventully exploded nd vnished.
Blck holes the mss of the sun would tke 1064 yers to explode. For blck hole
s the
mss of  glxy those fireworks would light up spce-time 1098 yers from now.
40Aginst the Fll of Night
Will there be nything or nyone round to see these quntum fireworks?
Dr. Dyson rgued in his 1979 pper tht life nd intelligence could survive the
desert of
drkness nd cold in  universe tht ws expnding infinitely but ever more slow
ly by
dopting ever slower nd cooler forms of existence. Intelligence, could reside,
exmple, in the pttern of electriclly chrged dust grins in n interstellr c
sitution described in the 1957 science fiction novel "The Blck Cloud," by the
stronomer Sir Fred Hoyle, who died in August.
As n orgnism like the blck cloud cooled, he rgued, it would think more slowl
y, but it
would lwys metbolize energy even more slowly, so its ppetite would lwys be
thn its output. In fct, Dr. Dyson concluded, by mking the mount of energy ex
per thought smller nd smller the cloud could hve n infinite number of thoug
while consuming only  finite mount of energy.
But there ws  hitch. Even just thinking requires energy nd genertes het, wh
ich is
why computers hve fns. Dr. Dyson suggested tht cretures would hve to stop
thinking nd hibernte periodiclly to rdite wy their het.
In n ccelerting universe, however, there is n dditionl source of het tht
cnnot be
gotten rid of. The sme clcultions tht predict blck holes should explode ls
o predict
tht in n ccelerting universe spce should be filled with so-clled Hwking r
In effect, the horizon -- the frthest distnce we cn see -- looks mthemticl
ly like the
surfce of  blck hole. The mount of this rdition is expected to be incredib
ly smll -corresponding to  frction of  billionth of  billionth of  billionth of  de
gree bove
bsolute zero, but tht is enough to doom sentient life.
"The Hwking rdition kills us becuse it gives  minimum temperture below whi
you cnnot cool nything," sid Dr. Kruss. Once n orgnism cools to tht tempe
he explined, it would dissipte energy t some fixed rte. "Since there is  fi
nite totl
energy, this mens  finite lifetime."
Infinity on Tril
Although Dr. Dyson grees with this gloomy view of life in n ccelerting unive
rse, he
nd Dr. Kruss nd Dr. Strkmn re still rguing bout whether life is lso doo
med in 
universe tht is not ccelerting, but just expnding nd getting slower nd col

Quntum theory, the Cse Western uthors point out, limits how finely the energy
new thoughts cn be shved. The theory decrees tht energy is emitted nd bsorb
ed in
tiny indivisible lumps clled "qunt." Any computtion must spend t lest this
energy out of  limited supply. Ech new thought is  step down n energy ldder
41finite number of steps. "So you cn only hve  finite number of thoughts," s
id Dr.
"If you wnt to stre t your nvel nd not think ny new thoughts, you won't di
energy, " he explined. But tht would be  boring wy to spend eternity. If lif
e is to
involve more thn the eternl reshuffling of the sme dt, he nd Dr. Strkmn
sy, it
cnnot be eternl.
Dr. Dyson, however, sys this rgument pplies only to so-clled digitl life, i
n which
there is  fixed number of quntum sttes. Cretures like the blck cloud, which
grow long with the universe, he sid, would hve n incresing number of quntu
sttes, nd so there would lwys be more rungs of the ldder to step down. So t
bottom need never be reched nd life nd thought could go on indefinitely.
But nobody knows whether such  life form cn exist, sid Dr. Kruss.
Compred with the sight of the World Trde Center towers collpsing or the pligh
t of 
sick child, this future extinction my seem  remote concern. Dr. Alln Sndge,
stronomer t Crnegie Observtories in Psden, Clif., who hs spent his life
investigting the expnsion nd fte of the universe, sid: "Life on this erth
is going to
vnish in 4.5 billion yers. I wouldn't get hung up on the fct tht the lights
re ll going
out in 30 billion yers."
Dr. Dyson sid he ws still n optimist. It is too soon to strt pnicking, he c
ounseled in
n e-mil messge. The observtions could be wrong.
"At present ll possibilities re open," he wrote. "The recent observtions re
not becuse they nswer the big questions bout the history of the universe, but
they give us new tools with which to explore the history."
Even in n ccelerting universe, Dr. Dyson sid, humns or their descendnts mi
ght one
dy be ble to rerrnge the glxies nd sve more of them from disppering. A
glimmer of hope comes from the dedly nd chilling Hwking rdition itself, si
d Dr.
Rphel Bousso, from the Institute of Theoreticl Physics t the University of C
t Snt Brbr. Since tht rdition is produced by unpredictble quntum fluc
he pointed out, if you wit long enough nything cn pper in it, even  new un
"Sooner or lter one of those quntum fluctutions will look like  Big Bng," h
e sid.

In tht cse there is the possibility of  future, if not for us, t lest for s
omething or
somebody. In the fullness of time, fter ll, physics teches tht the improbbl
e nd even
the seemingly impossible cn become the inevitble. Nture is not done with us y
et, nor,
s Dr. Dyson indictes, re we necessrily done with nture.
We ll die, nd it is up to us to decide who nd wht to love, but, s Dr. Weinb
pointed out in  recent rticle in The New York Review of Books, there is  cert
nobility in tht prospect.
42"Though wre tht there is nothing in the universe tht suggests ny purpose
ure brings bout n increse of the thickness nd  shortening of the surfce, while, on the
other hnd, tension leds to splitting of the continentl blocks. The individul stges of perceived s
mountin formtion comprised continul processes
of splitting nd compression, whereby the originl
Silic crust (for which Wegener ssumed  thickness
of bout 3035 km) grdully decresed in surfce
re, split into seprte pieces, nd incresed in thickness. Along with the movement of the continentl
blocks,  hypothesized universl ocen (Pnthlss)
egan to divide into a shallow sea and a deep sea.
Volcanism, for Wegener, was mainly related to the
continental fronts. Areas where tension prevailed,
such as the Atlantic Ocean, and also opening faults,
seemed to e relatively poor in volcanoes as compared
with areas such as the Pacific Ocean, where pressure
was increasing. The fronts of moving locks made
conditions more favorale to volcanism than did the
acks. Nevertheless, Wegener wondered whether the
mid-Atlantic ridge might e considered as a zone
where, with the continuing expansion of the Atlantic,
the floor was continuously reaking up, making room
for fresh, relatively fluid and high-temperature Sima
from elow! Moreover, increased volcanic activity in
some periods of Earth history might e due to large
displacements (as, for instance, during the Tertiary).
Trench faults (Graenru che), i.e., rift valleys,
acquired new meaning as representing the beginnings
of new continental separations. Gravity measurements had shown that beneath such lines lay material
of greater density, compared to that on either side.
Therefore, these lines could be seen as incipient fissures within the continental blocks (into which the
denser Sima was rising according to the principle of
isostasy). The best examples of such separations were
provided by the East African trenches and their continuation through the Red Sea. At the majority of the
trenches, the measurable mass deficit was not compensated by greater density of the matter beneath it.
Thus, the trenches must be youthful disruptions of a
continental block.
Wegeners theory of mountain formation was further supported y the fact that the folding of the
Andes seems to have een essentially simultaneous
to the opening of the Atlantic Ocean. The American

locks, during their westward drifting, had encountered resistance at the presumaly very old and relatively rigid floor of the Pacific Ocean. Thus, the
extended shelf, with its mighty sediments, forming
the western order of the continental lock, was
compressed to a range of fold mountains. For the
Tertiary folds of the Himalayas, Wegener assumed
that lower India had formed an extended peninsula
prior to compression, the southern end of which lay
next to that of South Africa. The folds had een
produced y impact of the Indian sucontinent and
the main mass of Asia.
Geological and Palaeontological
The palaeontological evidence indicating a former
connection etween the organic components of different continents had already given rise to the doctrine
of former land ridges. Among the most striking
findings were the distriutions of the Glossopteris
flora on the southern continents and the occurrence
of Mesosaurus at the turn of the Permian and the
Caroniferous exclusively in south-eastern South
America and the western parts of Africa; oth of
these discoveries suggested a former connection of
the two continents. Using these relationships also
allowed calculations of when the continents were
separated (either y horizontal displacements or y
sinking of the land ridges). South America and Africa
had een connected during the Mesozoic, ut were
separated at the end of the Eocene or Early Oligocene.
The connection etween Europe and North America
seemed to have een maintained during the older
Tertiary period, ut separation occurred in the
Miocene, although it might have continued in the far
north (over Scandinavia and Greenland) into the
Pleistocene. The connection of Lower India with
southern Africa, which Wegener had postulated
ased on his ideas on the formation of the Himalayan
range, was also confirmed y palaeontological evidence. Zoogeographers had long assumed a former
elongated IndianMadagascan peninsula (called
Lemuria), separated from the African lock y the
Mozamique Channel.
The zoogeographic concept of Lemuria had given
rise to Suess notion of a great southern continent, Gondwana, comprising parts of South America,
Africa, Lower India, Australia, and Antarctica.
Assuming the unchanged positions of its present-day250 FAMOUS GEOLOGISTS/Wegener
relics, however, required ascriing a huge extent to
this continent. Wegener, y contrast, proposed a
much reduced primeval continent, Pangaea. In the
Permian, i.e., until some 300 Ma ago, all the continents were supposedly joined in one land mass
extending from pole to pole. During the Triassic,
aout 200 Ma ago, Pangaea egan to reak up and
the newly emerging continents started moving into
their current positions. In the Jurassic, there were
few remaining connections except at the northern
and southern ends. Just as northern Europe and
North America remained connected until the older

Tertiary period, a connection of the southern continents seems to have persisted, running from the
southern coast of Australia over Antarctica to South
America. Later, the Antarctic lock, like the
South American lock in the Tertiary, moved over
from South Africa towards the side of the Pacific
Ocean. Only in the Quaternary period, then, did the
Australian lock ecome detached (Figure 4).
For geological and tectonic evidence, Wegener
referred particularly to Suess magnum opus, pulished in three volumes during 18851909, Das
Antlitz der Erde (The Face of the Earth). Considering
the tectonic relations, Europe/Africa and oth Americas seemed to represent the edges of an immense
expanded fissure. In the north, for instance, the Greenland massif was matched y Scandinavia, oth consisting of gneiss, and the less mountainous North
America corresponded to the likewise less mountainous Europe. The most striking example, however, was
the Caroniferous mountain range, called the Armorican mountains (Suess transatlantic Altaides),
which made the coalfields of North America appear
to e the direct continuation of the European ones.
Wegeners theory of mountain formation was also
confirmed y remarkale differences etween the
Atlantic and the Pacific hemispheres, such as the
distinction etween Pacific and Atlantic types of
coasts (marginal chains and ocean trenches in front
Figure 4 Wegeners reconstruction of the separation of the continents from the pri
meval Pangaea, from his 1926 paper Pala ogeo
graphische Darstellung der Theorie der Kontinentalverschiebungen , showing the r
elative positions of the continents during the Upper
Carboniferous (Jung Karbon), Eocene (Eozan), and Lower Quaternary (Alt Quartar)
(in two different projections). Cross hatching
represents deep seas, dotted regions represent shallow seas; rivers, recent coas
tlines, and outlines are shown only for orientation.FAMOUS GEOLOGISTS/Wegener 25
of the Pacific coasts, as contrasted to the wild, irregular ria Atlantic coastlines). There were also differences in the volcanic lavas of the two hemispheres, as
emphasized y the Vienna petrographer Friedrich
Becke (18551931) and others. The Atlantic lavas
contained a greater proportion of sodium, whereas
calcium and magnesium prevailed in the Pacific
lavas. Such differences were intelligile according to
the assumptions of continental movements. The
opening of the Atlantic was matched y the general
pressing of the continents against the region of the
Pacific Ocean: pressure and compression prevailed
at the coasts of the latter whereas tension and
splitting occurred at the latter.
Traces of glaciation during the Permian (ground moraines lying on scratched edrock) were to e found
on the southern continents, e.g., in East India and
Australia. If the present-day arrangement of the land
masses had prevailed at that time, this Permian ice
age would have required an icecap of seemingly impossile size. And the north pole would have een in
Mexico, where no trace of glaciation during that

period was recorded. Following the idea of horizontal

displacements, however, all regions sujected to
glaciation came together concentric to the southern
margin of Africa. And one had only to place the south
pole in this much reduced glaciated area to give the
Permian ice age a much more plausile form.
Wegener had discussed these palaeoclimatological
features since 1912. In 1924, he gave a detailed
description of the climatological changes from the
Caroniferous through to recent times, following the
traces of glaciations, swamps, and deserts, i.e., moraines, coal, salt, and gypsum, throughout Earths
history (Figure 5). In reconstructing the respective
polar shifts, Wegener emphasized that they oviously
took place along with the great displacements of the
continental locks. In particular, there was temporal
coincidence of the est confirmed polar shift, in the
Tertiary, and the opening of the Atlantic (Figure 6).
Movement of the poles since the Pleistocene might also
e related to the final separations of the continents in
the north and the south.
Motive Forces
Wegener was very cautious aout the forces that might
have caused continental displacements. First, it was
necessary to demonstrate the reality and the manner of
the displacements efore indulging in the hope of finding their cause. Nevertheless, he tentatively suggested
two candidates: centrifugal forces caused y the rotation of Earth and tidal-type waves within Earth, generated y the gravitational pull of the sun and the
moon. In the 1929 revision of Wegeners theory in
Figure 5 Wegener thought continental drift was the key to the climatic changes d
uring Earths history. This map, pulished in the
1924 ook y Koppen and Wegener, Die Klimate der geologischen Vorzeit , shows tr
aces of glaciation, swamps, and deserts for the
Caroniferous. E, Traces of glaciation; K, coal; S, salt; G, gypsum; W, desert s
andstone. Dotted regions indicate arid areas, dashed
lines indicate the positions (i.e., the pathways) of the poles, and the old cur
ved line indicates the respective position of the equator.252 FAMOUS GEOLOGISTS/
Figure 6 Map pulished in the 1924 ook y Koppen and Wegener, Die Klimate der g
eologischen Vorzeit , showing polar shifts (dashed
lines) from the Caroniferous to recent, related to the African tale (left, sou
th pole; right, north pole). Bold lines outline the continental
locks; hatched lines represent the Caroniferous (Karon) period. Perm, Permian
; Jura, Jurassic; Trias, Triassic; Kreide, Cret
aceous; Eozan, Eocene; Miozan, Miocene; Beginn des Quartar, eginning of the Qua
Die Klimate, he also mentioned convection currents
within the Sima; these had een first discussed as a
cause of mountain formation y the Vienna geologist
Otto Ampferer (18751947) in 1906.
Wegener also endeavoured to calculate the recent
velocity of the relative motion of the continents, though he was well aware that these values must e
quite uncertain. In his 1912 paper, comparing various
longitude determinations for Greenland, he had
deduced an increase of the distance to Europe of
11 m year 1 . Referring to the lengths of transatlantic
cales, he suggested that North America was drifting

away from Europe at aout 4 m year 1 .

From Continental Drift to
Plate Tectonics
The theory of continental drift was long rejected y
the majority of geologists. Among Wegeners few
followers were the South African Alexander Du Toit
(18781948), for whom continental drift provided
the est explanation of the close similarities etween
the strata and fossils of Africa and South America,
and the Swiss geologist E mile Argand (18791940),
who saw continental collisions as the only means of
producing the folded and uckled strata he had oserved in the Alps (see Famous Geologists: Du Toit).
Nevertheless, Wegeners explanation of the PermoCaroniferous ice age impressed even his critics.
Wegeners reputation as a meteorologist and a
polar explorer contriuted to keeping his theory
alive. His work was immediately rememered when,
around 1960, surprising data were otained from the
ocean floor: palaeomagnetic patterns alongside the
mid-ocean ridges clearly suggested the spreading of
the seafloor. Within aout two decades, Wegeners
principle of horizontal displacements of parts of
Earths crust ecame almost universally accepted, although, ironically, the process still lacked a consensus
as to its causes, though convection currents in the
internal mantle are most commonly advocated.
It should e noted that Wegeners original concept
differed from modern plate tectonics in essential
points, particularly with regard to the Sial and the
Sima. According to modern theory, the (Sialic) continents do not plough through the (oceanic) Sima.
Instead, oth continents and ocean floor are regarded
as forming solid plates, floating on the asthenosphere, which, due to tremendous heat and pressure,
ehaves like an extremely viscous liquid (as Wegener
had thought the Sima did). Therefore, the older
term continental drift, still often used today, is not
quite appropriate for the modern concept. Notwithstanding these differences, Wegeners asic ideas
remain sound, and the lines of evidence that he
used to support his theory are still valid. He first
envisaged a dynamic Earth, connecting its major
features and various geological processes continentalFLUID INCLUSIONS 253
movements, folded mountain ranges, rift systems,
earthquakes, volcanism, ocean transgressions, palaeoclimatological changes, etc. on a gloal scale. In
this sense, Wegeners theory was a true forerunner of
plate tectonics.
See Also
Africa: Rift Valley. Famous Geologists: Du Toit; Suess.
Gondwanaland and Gondwana. History of Geology
From 1900 To 1962. History of Geology Since 1962.
Palaeoclimates. Pangaea. Plate Tectonics. Tectonics:
Mid-Ocean Ridges; Mountain Building and Orogeny.
Further Reading
Carozzi AV (1985) The reaction of continental Europe
to Wegeners theory of continental drift. Earth Sciences
History 4: 122 137.
Fritscher B (2002) Alfred Wegeners The origin of contin

ents, 1912. Episodes 25: 100 106.

Jacoy WR (2001) Translation of Die Entstehung der Kon
tinente, Dr Alfred Wegener, Petermanns Geographische
Mitteilungen, 58 (1912). Journal of Geodynamics 32:
29 63.
Ko ppen V and Wegener A (1924) Die Klimate der geolo
gischen Vorzeit. Berlin: Borntra ger.
Lu decke C (1994) Stratigraphische Methode der Rekon
struktion von Expeditionsergebnissen am Beispiel des
Todes von Alfred Wegener wa hrend der Gro nlandexpedi
tion (1930 31). In: Fritscher B and Brey G (eds.) Cosmo
graphica et Geographica: Festschrift fu r Heribert M.
Nobis zum 70. Geburtstag, Algorismus, vol. 13,
pp. 347 367. Munich: Institut fu r Geschichte der
Oreskes N (1999) The Rejection of Continental Drift:
Theory and Method in American Earth Science. New
York and Oxford: Oxford University Press.
Runcorn SK (ed.) (1966) Continental Drift. New York and
London: Academic Press.
Schwarzbach M (1986) Alfred Wegener: The Father
of Continental Drift. Madison. WI: Science Tech
Sengo r AMC (1991) Timing of orogenic events: a persistent
geological controversy. In: Mu ller DW, McKenzie JA,
and Weissert H (eds.) Controversies in Modern Geology:
Evolution of Geological Theories in Sedimentology,
Earth History and Tectonics, pp. 403 473. London:
Academic Press.
Wegener A (1912) Die Entstehung der Kontinente.
Petermanns Mitteilungen aus Justus Perthes Geogra
phischer Anstalt 58: 185 195, 253 256, 305 309.
Wegener A (1926) Pala ogeographische Darstellung der
Theorie der Kontinentalverschiebungen. In: Dacque E
(ed.) Pala ogeographie, pp. 171 189. Leipzig and Wien:
F Deuticke.
Wegener A (1971) The Origin of Continents and Oceans.
(Translation from the 4th revised German edition by
J Biram, with an introduction by BC King.) London:
Wegener A (1980) Die Entstehung der Kontinente und
Ozeane. (Reprint of the 1st and 4th editions, edited by
A Vogel.) Braunschweig: Vieweg.
Wegener E (1960) Alfred Wegener: Tagebu cher, Briefe,
Erinnerungen. Wiesbaden: Brockhaus.
Wutzke U (1998) Kommentiertes Verzeichnis der schriftli
chen Dokumente seines Lebens und Wirkens, Berichte
zur Polarforschung 288. Bremerhaven: Alfred Wegener
Institut fu r Polar und Meeresforschung.
A H Rankin, Kingston University,
Kingston-upon-Thames, UK
2005, Elsevier Ltd. All Rights Reserved.
Fluid inclusions are small droplets of fluid that have
been trapped within crystals either during primary
growth from solution or at some later stage, usually
as a result of recrystallization along healed microfractures. They are ubiquitous in both naturally
occurring minerals and in laboratory-grown crystals.

To the chemist or materials scientist, these gross

defects cause endless obstacles in their quest to grow
near-perfect crystals. However, to the geologist, they
provide a unique fossil record of the various
fluids responsible for the formation and evolution of
rocks and minerals throughout the history of the
Despite their small size (usually less than 20 mm),
their chemical composition and physical properties
can be readily determined, and the data may be used
to estimate the temperatures, pressures, and physicochemical nature of the fluid at the time of trapping.
This information has made an immense contribution
to the development of modern theories of ore genesis,
petrogenesis, diagenesis, and petroleum migration
and accumulation, and to our understanding of the
importance of the fluid phase in a wide range of
geological processes.254 FLUID INCLUSIONS
Occurrence and General
Formation and Genetic Classification of Fluid
Small changes in the chemical or physical properties
of fluids near to a growing crystal face can lead to
perturbations in the stability of crystal growth and the
development of gross defects, manifested as embayments, along crystal faces. These embayments will
seal over during a period of greater stability, trapping
a portion of fluids to form primary (P) fluid inclusions. In many instances, the trapped fluid will e
homogeneous at the time of trapping. In others,
where immiscile fluids are present or where mechanical entrapment of other coexisting crystalline phases
has occurred, trapping will e heterogeneous.
At some stage after primary growth, secondary (S)
fluid inclusions can form from later fluids, particularly
as a result of recrystallization along microfractures.
The chemical and physical properties of these inclusions may e very different from those of the earlier
mineral-forming fluids. However, if fracturing and
rehealing take place during primary growth, the fluids
may e indistinguishale, and the terms pseudosecondary or primarysecondary (PS) appropriately
descrie such inclusions. A schematic representation
of this genetic classification of inclusions is shown in
Figure 1.
For most geological applications, it is necessary to
estalish whether the inclusions are primary, secondary, or pseudosecondary, and also whether heterogeneous trapping has occurred. Heterogeneous trapping
may e recognized y the variale proportions of
liquids and solids in a single group or generation of
inclusions. Various criteria may e used to distinguish
etween P, PS, and S inclusions, ut these may e
difficult to apply and it may e difficult to identify
primary inclusions in many samples.
Figure 1 Schematic representation of the distriution of pri
mary (P), secondary (S), and pseudosecondary (PS) fluid inclu
sions in a quartz crystal. Modified from Rankin AH (1989) Fluid
inclusions. Geology Today 5: 21 24.

in turid or translucent minerals, such as feldspar.

Quartz is usually the preferred host.
Size and Shape of Inclusions
Choice of Material for Study
The successful application of fluid inclusion studies
depends partly on serendipity and partly on the type
and quality of material availale for study. Due to
their small size, oservations on fluid inclusions are
carried out under a microscope using polished wafers
around 12 mm thick. In most cases, clear, transparent minerals are needed, ut it is also possile to study
inclusions in some deeply coloured, semi-transparent
minerals in very thin (<50 mm) polished sections.
Care must e taken with soft, easily cleaved minerals,
such as calcite and fluorite, ecause of the possiility
of leakage during sample preparation or analysis.
Fluid inclusions are particularly difficult to oserve
Fluid inclusions seldom exceed 1 mm in size; most are
less than 20 mm, and those greater than 1 cm are
exceptionally rare and usually regarded as museum
specimens. Fluid inclusions display a variety of
shapes. They may e flattened and irregular, rounded,
or regular with three-dimensional negative crystal
shapes mimicking the crystal symmetry of the host
crystal (Figure 2).
Phases Present at Room Temperature
Fluid inclusions contain varying proportions of liquid
(L), solid (S) and gas (G) depending on the composition (X), temperature (T), pressure (P), and volume
(V) of the enclosed fluid at the time of entrapment.
humanity," he wrote, "one way that we can find a purpose is to study the univers
e y the
methods of science, without consoling ourselves with fairy tal
Niels Bohr, a Danish physicist and leader of this revolution, once said that a p
erson who
was not shocked y quantum theory did not understand it.
This week, some 700 physicists and historians are gathering in Berlin, where Pla
started it all 100 years ago, to celerate a theory whose meaning they still do
understand ut that is the foundation of modern science. Quantum effects are now
7invoked to explain everything from the periodic tale of the elements to the ex
istence of
the universe itself.
Fortunes have een made on quantum weirdness, as it is sometimes called. Transis
and computer chips and lasers run on it. So do CAT scans and PET scans and M.R.I
machines. Some computer scientists call it the future of computing, while some p
say that computing is the future of quantum theory.
"If everything we understand aout the atom stopped working," said Leon Lederman
former director of the Fermi National Accelerator Laoratory, "the G.N.P. would
go to
The revolution had an inauspicious start. Planck first regarded the quantum as a
ookkeeping device with no physical meaning. In 1905, Alert Einstein, then a pa
clerk in Switzerland, took it more seriously. He pointed out that light itself 

ehaved in
some respects as if it were composed of little energy undles he called lichtqua
nten. (A
few months later Einstein invented relativity.)
He spent the next decade wondering how to reconcile these quanta with the tradit
electromagnetic wave theory of light. "On quantum theory I use up more rain gre
than on relativity," he told a friend.
The next great quantum step was taken y Bohr. In 1913, he set forth a model of
the atom
as a miniature solar system in which the electrons were limited to specific ori
ts around
the nucleus. The model explained why atoms did not just collapse -- the lowest o
rit was
still some slight distance from the nucleus. It also explained why different ele
emitted light at characteristic wavelengths -- the orits were like rungs on a l
adder and
those wavelengths corresponded to the energy released or asored y an electron
it jumped etween rungs.
But it did not explain why only some orits were permitted, or where the electro
n was
when it jumped etween orits. Einstein praised Bohr s theory as "musicality in
sphere of thought," ut told him later, "If all this is true, then it means the
end of
While Bohr s theory worked for hydrogen, the simplest atom, it ogged down when
theorists tried to calculate the spectrum of igger atoms. "The whole system of
of physics must e reconstructed from the ground up," Max Born, a physicist at
Gottingen University, wrote in 1923. He termed the as-yet-unorn new physics "qu
Boy s Mechanics
The new physics was orn in a paroxysm of deate and discovery from 1925 to 1928
has een called the second scientific revolution. Wolfgang Pauli, one of its rin
called it "oy s mechanics," ecause many of the physicists, including himself,
then 25,
8Werner Heisenerg, 24, Paul Dirac, 23, Enrico Fermi, 23, and Pascual Jordan, 23
, were
so young when it egan.
Bohr, who turned 40 in 1925, was their father-confessor and philosopher king. Hi
s new
institute for theoretical physics in Copenhagen ecame the center of European sc
The decisive moment came in the fall of 1925 when Heisenerg, who had just retur
ned to
Gottingen University after a year in Copenhagen, suggested that physicists stop
trying to
visualize the inside of the atom and instead ase physics exclusively on what ca
n e seen
and measured. In his "matrix mechanics," various properties of suatomic particl
es could
e computed -- ut, disturingly, the answers depended on the order of the calcu

In fact, according to the uncertainty principle, which Heisenerg enunciated two

later, it was impossile to know oth the position and velocity of a particle at
once. The
act of measuring one necessarily distured the other.
Physicists uncomfortale with Heisenerg s astract mathematics took up with a
friendlier version of quantum mechanics ased on the familiar mathematics of wav
es. In
1923, the Frenchman Louis de Broglie had asked in his doctoral thesis, if light
could e a
particle, then why couldn t particles e waves?
BALTIMORE, April 5 -- A gasp went through the auditorium of the Space Telescope
Science Institute on Wednesday when Dr. Adam Riess, a young astronomer from the
institute, put the last mark on his so-called Hule diagram, a plot of the rig
htness and
speed of distant ojects that astronomers use to divine the history of the unive
The Darth Vaders of astronomy had gathered here to take stock of their expanding
increasingly dark realm. Once upon a time astronomy was aout what could e seen
the sky, aout jewel-like lights that moved in eternally recurring patterns and
the soft
glow of galaxies and comets.
Now it is aout what cannot e seen. In the last few decades astronomers have ha
d to
confront the possiility that stars and galaxies -- not to mention the creatures
that inhait
them -- are arely more than flecks of froth on a stormy sea of dark matter.
Now Dr. Riess was presenting his colleagues with evidence, ased on oservations
of a
star that exploded 11 illion years ago, that the universe -- dark matter and al
l -- was
eing lown apart under the influence of a mysterious antigravitational force kn
own only
as "dark energy."
"We are doing astronomy of the invisile," admitted Dr. Mario Livio, a theorist
at the
Space Telescope institute, who had organized the meeting, called "The Dark Unive
Matter, Energy, and Gravity" last fall.
As it turned out, the meeting coincided with a NASA news conference announcing t
reakthrough discovery y Dr. Riess and his colleagues and thus was dominated y
discussions of new telescopes in space and new dimensions in the universe as
astronomers grappled with the meaning of dark energy and how to take its measure
Now physicists, some of whom have een reluctant to take acceleration of the uni
seriously, will have to explain what this dark energy is. "Those numers are ala
and apparently true," said Dr. Michael Dine, a theoretical physicist from the Un
iversity of
California at Santa Cruz. He descried his colleagues as now working "franticall
y" to
find an explanation.
19On one level, the universe, with all of its dark aggage, seems to make sense.
The total
amount of matter and energy seems to e just enough to guarantee that the large-

geometry of space-time is "flat," or Euclidean, a result that cosmologists have
considered to e the most desirale and aesthetic. On the other hand, the detail
reakdown of the constituents of the cosmos is, as Dr. Livio says, "ugly" -- 65
dark matter, 30 percent dark matter of unknown nature and only 5 percent stars,
gas and
"We live in a preposterous universe," said Dr. Michael Turner, an astrophysicist
at the
University of Chicago. "Dark energy. Who ordered that?"
Of course, it was Einstein who originally ordered dark energy when he inserted a
factor called the cosmological constant into his gravitational equations descri
ing the
universe. Lamda, as it is known, after the Greek letter, represented a sort of
repulsion associated with space itself that kept the cosmos from collapsing of i
ts own
weight. Einstein aandoned the cosmological constant when it was discovered that
universe was expanding, and resisted efforts to ring it ack, once referring to
it as his
iggest lunder.
But he couldn t keep it out forever. In 1998 two competing teams of astronomers
trying to
measure how the expansion of the universe was slowing down ecause of cosmic gra
found that the universe was actually speeding up, as if the galaxies were eing
apart y a force -- dued, in the spirit of the times, "dark energy."
"This was a very strange result," recalled Dr. Riess, who was a memer of one of
teams. "It was the opposite of what we thought we were doing." Was this Einstein
s old
cosmological constant, something even weirder or a mistake?
The effect had showed up as an unexpected dimness on the part of certain explodi
ng stars
known as supernovae that the astronomers were using as so-called standard candle
ojects whose distance could e gauged from their apparent rightness. The astro
deduced that these stars were farther away than they should have een in an even
expanding universe, and that therefore the expansion was actually accelerating.
But dust or chemical changes over the eons in the stars could also have dimmed t
supernovae. The cleanest test of dark energy and the acceleration hypothesis, Dr
. Riess
explained, would e to find supernovae even farther out and ack into the past,
or more ack to the Big Bang itself. Because it is space itself that provides th
e repulsive
push, according to Einstein s equations, that push should start out small when t
universe is small and grow as the universe expands. Cosmic acceleration would on
ly kick
in when lamda s push got ig enough to dominate the gravity of ordinary matter

energy in the universe, aout five or six illion years ago. Before then the uni
verse would
have een slowing down, like a Mark McGwire last that has not yet reached the t
op of
its trajectory, and a supernova glimpsed at that great distance would appear rel
righter than it should. If dust or chemical evolution were responsile, such di
stant stars
should appear relatively even dimmer.
20By chance the Hule Space Telescope had oserved a supernova in late 1997 and
1998 that proved to e 11 illion light-years away -- the most distant yet seen.
On Dr.
Riess s Hule diagram it appeared twice as right as it should.
"Extraordinary claims require extraordinary evidence -- I hope the I.R.S. doesn
t say that
to you," Dr. Riess told his audience, ut, he concluded, "the cosmological const
ant looks
good for this supernova."
Dr. Livio said, "A year ago proaly a large fraction of the people in this room
would not
have elieved it."
But there were more complicated explanations, forms of dark energy other than th
cosmological constant on physicists drawing oards, as well as the possiility
astronomers were still eing fooled. To explicate the nature of the dark energy,
astronomers need to oserve more supernovae as far ack as 11 illion years ago,
cover the time when the universe egan to accelerate.
"How fast did it go from deceleration to acceleration?" asked Dr. Riess. Answeri
ng such
questions could help astronomers determine how hard the dark energy is pushing o
n the
universe compared with the predictions for the cosmological constant. A fast tur
he said, "egins to tell you there is a lot of oomph for a given amount of it.
"The cosmological constant is the enchmark oomph," he said.
To find those supernovae so far out cosmologists will have to go to space, said
Dr. Saul
Perlmutter, a physicist at the University of California s Lawrence Berkeley Nati
Laoratory and a veteran dark energy hunter.
On the ground, the supernova researchers have to employ a wide network of people
telescopes to detect the explosions, diagnose their type and then to watch them
fade. Dr.
Perlmutter descried an oriting telescope that would perform all three function
s. The
Supernova/Acceleration Proe, or SNAP, would comine an 80-inch diameter mirror
(only aout 16 percent smaller than the Hule), a giant electronic camera with
a illion
pixels and a special spectroscope.
If all goes well, Dr. Perlmutter said, the telescope could e launched in 2008.
In three
years of operation, he estimated, SNAP could harvest aout 2,000 supernovae. To
distinguish the different ideas aout dark energy, oservations would have to e

to the level of 1 or 2 percent uncertainty .

"We re all excited," he said.
So are the physicists. Their list of suspects egins with Einstein s cosmologica
l constant,
ut therein lies a prolem. Aout the time that Einstein was aandoning it, quan
mechanics -- the set of rules that govern the suatomic realm -- was estalishin
g a
21theoretical foundation for the cosmological constant. According to quantum the
empty space should e foaming with temporary particles and their cumulative ener
would outweigh the matter in the universe, including dark matter, y 120 orders
magnitude -- that is, a factor of 10 followed y 119 zeros. At that level, the f
orce of the
vacuum would either have crumpled the universe or lown it apart efore even an
had the chance to form.
The fact that the universe is in fact puttering along rather gently suggests tha
t there is
something fundamental aout physics and the universe that physicists still don t
Dr. Steven Weinerg, the Noel Prize-winning particle theorist at the University
Texas, has called the cosmological constant "the one in our throat." If the dar
k energy
really is Einstein s cosmological constant, then physicists have to answer quest
ions like
why it is so small -- roughly comparale, in fact, to the density of matter in o
ur own
Lacking an answer so far, even from string theory, the mathematically daunting c
theory of everything, some theorists have resorted to a controversial and somewh
philosophical notion called the anthropic principle, which, in effect, says that
need to factor in their own existence when they think aout the universe. Out of
all the
possile universes that can e imagined, this line of thinking goes, the only on
es in which
humans can find themselves is one that is conducive to human life.
That means, Dr. Weinerg has pointed out, that the cosmological constant has to
e small
enough to allow time for galaxies and stars to condense from the primordial fog
efore it
takes over and starts lowing the universe apart.
Dr. Alex Vilenkin of Tufts University in Massachusetts pointed out to the Dark U
audience that the universe was at its peak in making stars aout five illion or
six illion
years ago, just aout the time that dark energy and the matter density would hav
e een
equal. Our own sun, some 4.5 illion years old, was on the tale end of that wave
, and now
here we are. "Oservers are where the galaxies are," said Dr. Vilenkin. "A typic
oserver will see a small cosmological constant."
Many physicists are uncomfortale with this line of reasoning, and they are seek

ing the
answer in different class of theories known as quintessence, after the Greek wor
d for the
fifth element. Modern physics, noted Dr. Paul Steinhardt, a theorist at Princeto
n, is
replete with mysterious energy fields that would exhiit negative gravity. The t
rick, Dr.
Steinhardt explained, is finding a field that would act like the dark energy wit
hout a lot of
fudging on the part of theorists.
"The oservations are forcing us to do this," he said. "Dark energy is an intere
prolem. Any solution is quite interesting."
One theory that captured the fancy of the astronomers in Baltimore was a modific
ation of
gravity recently proposed y three string theorists at New York University: Dr.
Dvali, Dr. Gregory Gaadadze and Dr. Massimo Porrati. In string theory -- so nam
22ecause it descries elementary particles as tiny virating strings -- the ord
inary world is
often envisioned as a three-dimensional island (a memrane, or "rane" in string
floating in a 10- or 11-dimensional space. Ordinary particles like electrons and
and forces like electromagnetism are confined to three dimensions, to the rane,
gravity is not.
As a result, Dr. Dvali suggested that gravity could only travel so far through c
space efore it leaked off into the extra dimensions, therey weakening itself.
To an
oserver in the traditional three dimensions it looks as if the universe is acce
lerating. The
cosmological constant, in effect, he said, is a kind of gravitational rane drai
n. "Gravity
fools itself," he said. "It sees itself as a cosmological constant."
Dr. Dvali s theory was welcomed y the astronomers as a sign that string theory
eginning to come down from its ivory tower of astraction and make useful, test
predictions aout the real world. (In another string contriution, Dr. Steinhard
t introduced
a new theory of the early universe, in which the Big Bang is set off y a pair o
f ranes
clashing together like cymals.)
Afterward Dr. Riess and Dr. Perlmutter pressed Dr. Dvali on what they would see
they looked out past the crossover point where gravity egan falling out of the
would the transition etween a decelerating universe and an accelerating one hap
more aruptly than in the case of the cosmological constant? Dr. Dvali said he h
adn t
done any calculations, ut he said it was his "nave guess" that the crossover wou
happen more smoothly than in a lamda world.
"I d love to see this guy do some Hule diagrams," Dr. Riess said.
Even if Dr. Dvali could e coaxed into providing a prediction, however, success

identifying the dark energy was not guaranteed to the astronomers. Calling himse
lf a
spokesman for the "cranky point of view," Dr. Steinhardt pointed out that the of
tproclaimed era of "precision cosmology" was ound to have its limits. Other
cosmological parameters, particularly the cosmic density of matter in the univer
se, were
not likely to e known well enough for even SNAP to untangle the models in which
quintessence varied over time. Worried that the overselling of SNAP could sap
astronomers will to come up with new ideas, he said, "We should try to make as
pronouncements as possile."
Dr. Turner refused to e swayed from his "irrational exuerance." Appealing to t
astronomers pride, he urged them to e amitious. "We have a chance to do funda
physics here," he said. "Let s see if we can crack this nut. Maye we ll fall on
our faces.
Maye cranky Paul is right.
"I still have a lot of youthful juices in my ody."
Copyright 2002 The New York Times Company
23Mysteries of the Universe
Before the Big Bang, There Was . . . What?
What was God doing efore he created the world? The philosopher and writer (and
saint) Augustine posed the question in his "Confessions" in the fourth century,
and then
came up with a strikingly modern answer: efore God created the world there was
time and thus no "efore." To paraphrase Gertrude Stein, there was no "then" the
Until recently no one could attend a lecture on astronomy and ask the modern ver
sion of
Augustine s question -- what happened efore the Big Bang? -- without receiving
same frustrating answer, courtesy of Alert Einstein s general theory of relativ
ity, which
descries how matter and energy end space and time.
If we imagine the universe shrinking ackward, like a film in reverse, the densi
ty of
matter and energy rises toward infinity as we approach the moment of origin. Smo
pours from the computer, and space and time themselves dissolve into a quantum "
"Our rulers and our clocks reak," explained Dr. Andrei Linde, a cosmologist at
University. "To ask what is efore this moment is a self-contradiction."
But lately, emoldened y progress in new theories that seek to unite Einstein s
realm with the unruly quantum rules that govern suatomic physics -- so-called q
gravity -- Dr. Linde and his colleagues have egun to edge their speculations cl
oser and
closer to the ultimate moment and, in some cases, eyond it.
Some theorists suggest that the Big Bang was not so much a irth as a transition
, a
"quantum leap" from some formless era of imaginary time, or from nothing at all.

others are exploring models in which cosmic history egins with a collision with
Inspired y de Broglie s ideas, the Austrian Erwin Schrodinger, then at the Univ
ersity of
Zurich and, at 38, himself older than the wunderkind, sequestered himself in the
resort of Arosa over the 1925 Christmas holidays with a mysterious woman friend
came ack with an equation that would ecome the yin to Heisenerg s yang.
In Schrodinger s equation, the electron was not a point or a tale, ut a mathem
entity called a wave function, which extended throughout space. According to Bor
n, this
wave represented the proaility of finding the electron at some particular plac
e. When it
was measured, the particle was usually in the most likely place, ut not guarant
eed to e,
even though the wave function itself could e calculated exactly.