The discovery of a lost ocean deep inside the Earth sounds like something
straight out of Jules Verne. Yet an American geologist believes he has indeed
managed this feat which, if true, could settle one of the hottest controversies
in modern geophysics.

About 210 million years ago, all the continents of the globe were collected
together in the giant supercontinent of Pangaea. It then began to split
apart, creating the Tethys Sea and dividing the continents of Laurasia in
the north and Gondwanaland in the south. Eventually, ammonites swam in the
waters and dinosaurs probably paddled around the shores. But the Tethys
Sea was not to last for long. It started closing up again and, by 45 million
years ago, Africa was colliding with Europe and India with Asia, creating
the great mountain ranges of the Alps and Himalayas.

Today, the Mediterranean is all that remains of the once expansive
ocean. But the floor of the Tethys may not have vanished without trace.
Michael Wysession of Washington University in St Louis has been analysing
seismic waves from earth tremors in the western Pacific. By combining data
from about 750 tremors, he has been able to build up one of the most detailed
deep cross sections of part of the Earth’s mantle.

He began by looking at a hot area beneath Indonesia that may represent
a plume of upwelling mantle material, but soon realised that the cold region
next to it was in the shape of a vast slab 250 kilometres thick, located
2 700 kilometres down in the mantle next to the outer core of the Earth.
Being cold and hard, this transmits the seismic waves more quickly than
the warm, softer material. Wysession believes that this is part of the lithosphere
that once made up the floor of the Tethys.

If he is correct, it adds to the increasing evidence for circulation
of the entire mantle over geological timescales. Previously, the majority
view was that the mantle was like a great double boiler with heat coming
all the way up from the core but with little or no chemical mixing across
a boundary about 650 kilometres down. If that boundary could not be crossed,
the ancient seafloor could not have reached the base of the mantle.

The truth may be a compromise. The 650-kilometre boundary clearly does
represent some sort of barrier. Other seismic profiles show the subduction,
or movement of one tectonic plate over another, of parts of the ocean floor
slowing as they approach the barrier, so that they spread out when they
reach it. In computer simulations last year, Paul Tackley of the California
Institute of Technology showed how such slabs might build up at the boundary
between the upper and lower mantle until they warmed sufficiently for mineral
phase changes to take place that would make them dense enough to continue
to sink all the way to the bottom of the mantle. If the cold slab that Wysession
has found is the lithosphere of the Tethys, it must have taken about 100
million years to get there. But even that is not the end. Slowly it will
heat up again, perhaps to the point where it starts to rise in a new mantle
plume. Then the same material may erupt from volcanoes along a new mid-ocean
ridge and another Tethys will open on the surface.

The way aspirin-like drugs work was deduced in 1971 by John Vale and
Salvador Moncada in Britain. Since then, the enzyme that aspirin blocks
has been isolated and sequenced. In fact, there are now known to be two
prostaglandin synthase enzymes which make prostaglandins, each with a different
function.

One prostaglandin, called thromboxin A2, makes blood platelets sticky
so that they coagulate, healing wounds but also causing atherosclerosis,
heart attacks and strokes. The second prostaglandin, produced after an injury,
causes inflammation and the pain of headaches and arthritis. The two forms
have about 60 per cent of their amino acid sequence in common, and aspirin
and aspirin-related drugs such as ibuprofen act on both. In fact, such
drugs inhibit the first form much more strongly than they do the one involved
in inflammation.

Very low doses are so good at reducing the stickiness of blood platelets
that they can help prevent heart attacks in people at risk. But controlling
the chronic pain of arthritis and other inflammatory diseases takes higher
doses which can disrupt the blood flow in the stomach lining, leading to
bleeding and ulcers. If a new drug could block the inflammation mechanism
without affecting the blood it would be good news for patients – and for
shareholders in the drugs company that developed it.

A team led by Michael Garavito of the University of Chicago has isolated
the enzyme prostaglandin H₂ synthase-1 from rams and crystallised
it, a difficult feat for an oily membrane protein (Nature, vol 367, p 243).
The researchers deduced the molecular structure from X-ray analysis. They
found that the molecule was mushroom-shaped with its base in the membrane
where the oily prostaglandin precursor, arachidonic acid, occurs. The enzyme
contains a groove which acts as a sort of assembly line for the prostaglandin.

So far, the researchers have determined only the structure of the first
form of the enzyme at relatively low resolution, but they are using computer
modelling to map out the details and hope soon to have a high resolution
structure for the second form.

There is little point in developing new drugs to block the first enzyme
because aspirin is cheap and effective and, when used for improving blood
flow, will work at doses far lower than those that cause ulcers. But an
anti-inflammatory drug that acts on the second enzyme alone is a holy grail
of pharmacology; a pain-killer without side effects – a sharp instrument
indeed.

The Antarctic ice sheet covers an area larger than Europe and contains
90 per cent of all the freshwater on the planet. If it were to melt entirely,
the sea level worldwide would rise by more than 60 metres. No one is anticipating
this, but a much smaller rise could still be devastating because much of
the world’s population lives in coastal areas.

Conference participants heard of past changes in the ice sheet, present
measurements and computer predictions for the future. Much of the past evidence
comes from ice cores; the Greenland Ice Core Project and the Vostok core
in Antarctica have reached back through a quarter of a million years of
ice.

Both show the same broad changes in temperature, snowfall and atmospheric
gases characteristic of the beginning and end of ice ages, but the Antarctic
does not show as many rapid changes in interglacial periods. This could
be because the continent is surrounded by a circumpolar ocean current whereas
Greenland is influenced strongly by currents that carry heat northwards
from the Atlantic. There is mounting evidence that the North Atlantic ‘conveyor
belt’ can suddenly become disrupted, sometimes by the influx of vast amounts
of fresh meltwater and at other times for unknown reasons.

Some of the latest ice core data come from new, shallow cores taken
in Antarctica and analysed in the past few weeks at the British Antarctic
Survey. They reveal temperature rises over the past few years that are far
greater than any in the past 400 years. ‘Modelling studies show that the
polar regions are going to be among the first to detect changes arising
from greenhouse warming,’ said David Peel of BAS, ‘but these changes are
amplifying even that.’

Many of the conference sessions concentrated on measurements of how
quickly the Antarctic ice is forming and melting today. It falls as snow
in the centre of the continent, compacts and flows outwards like a drop
of viscous liquid. Satellites are beginning to monitor the thickness of
the ice regularly, but have not been doing so long enough to reveal significant
trends.

One important factor is the speed at which the ice is flowing and what
slows or lubricates it. Richard Alley of Pennsylvania State University has
been studying ice streams that drain on to the Ross Ice Shelf in Western
Antarctica. Overall, the flows seem to be thinning and slowing but Alley
believes this is because the latest ice age ended 10 000 years ago and the
bottom of the ice sheet has only just started to react.

Ice will flow as long as its surface runs downhill, even if the ground
underneath runs uphill. Sometimes the ground topography underneath funnels
it into faster moving streams more like individual glaciers. The ice along
the edges of such streams is softer so that the streams can slip through
the sheet. The key factor in this process seems to be the lubrication underneath.
Local melting can produce a thin layer of water so that the ice moves easily.

Exactly what happens depends very much on the underlying geology. The
ice streams Alley is studying speed up as they thin out, but one of them
stopped suddenly about 130 years ago. Alley believes this was probably because
the water underneath drained away down a sloping rock surface and caused
the next flow to speed up – a process he calls ‘water piracy’.

Recent measurements of ice thickness have revealed what are probably
active volcanoes beneath the ice of Western Antarctica. They may well be
contributing water to the lubricating layer but no one knows whether or
not an eruption would cause a large part of the sheet to slip off. The only
other known example of volcanism under an ice sheet is on Iceland where
melting water runs out along natural pipes through the ice. But the pipes
are only a few kilometres long. The equivalents in Antarctica might need
to be a thousand kilometres long.

Where the ice floats out on to the sea to form the great ice shelves,
it can be up to 1500 metres thick, thinning to 300 or 400 metres before
the edge of open water. Glaciologists are keen to discover just how quickly
it is melting and breaking up, not only along the edge but underneath. To
do that, they have to drill a hole through the ice.

Keith Nichols and his colleagues at BAS have had two successful seasons
on the Ronne Ice Shelf. They use a long hose which squirts a high pressure
jet of hot water and melts about half a metre of ice a minute. But, once
started, it is vital nothing goes wrong or the hole will freeze up before
it is complete. Researchers lower slimline versions of standard oceanographic
instruments down the hole to measure conductivity and temperature and hence
salinity and density of the water. Then current meters are installed and
the holes left to freeze up around the cable so that data can be sent home
by satellite.

The picture emerging is of complex circulation in the cold, dark sea
under the ice. Seawater tends to freeze in winter around the edges of the
shelf, and water at freezing point sinks down under the ice towards the
coast. At that depth and pressure, the melting point of ice is as low as
-3 degree C and rapid melting takes place. The resulting cold, fresh water
sinks and flows out into the world’s oceans to form a layer of Antarctic
bottom water.

The edges of the ice shelves break up into icebergs in a process known
as calving. Olaf Orheim of the Norwegian Polar Research Institute has accumulated
data on more than 200 000 icebergs. The big ones are easy to monitor from
space – three the size of Cyprus broke off two years ago. But most icebergs
are too small to be seen by satellites and the Norwegians have to rely on
systematic observations from passing ships.

There was doubt at the conference over whether the results can allow
accurately for repeat observations of the same icebergs. But if estimates
are true, the ice shelves are breaking up far faster than expected.

None of the monitoring programmes have gone on long enough to make reliable
predictions. These are left to people using computers to model the ice,
the climate and the surrounding oceans. Bill Budd of the University of Tasmania
has run such a model hundreds of years into the future to see how the ice
might respond to global warming. He has found that the ice will take more
than 100 years to respond to today’s warming and that, although the ice
shelves will retreat and break up, more snow will fall inland, thickening
the ice and maintaining or even lowering sea level. Melting will not begin
to overtake thickening for well over 500 years.

The existence of haloes of dark matter was first deduced by Vera Rubin of
the Carnegie Institute in Washington DC from her observations of how stars
orbit in nearby galaxies. In 1982, Donald Linden-Bell of the University of
Cambridge and Douglas Lin of the University of California at Santa Cruz
suggested that such a halo might surround the Milky Way and reveal itself
by tugging on satellite galaxies such as the Large Magellanic Cloud (LMC)
and Small Magellanic Cloud (SMC).

The farther stars are from the centre of our Galaxy, the slower they are
orbiting. The Sun, for instance, completes an orbit of the Galaxy once every
200 million years. The LMC also appears to be orbiting, leaving behind a
trail of gas.

Lin and his colleagues set out to measure the motion of the LMC across the
sky – its ‘proper motion’. The faster it is going, the stronger must be the
Galaxy’s gravity in order to stop it flying off into space.

The astronomers studied the motions of 250 stars in the LMC and compared
photographs of the galaxy taken with the 4-metre telescope at Cerro Tololo
in Chile between 1974 and 1989. The motion they detected was so small it
was equivalent to watching grass grow in New York from London. This
corresponded to a speed of 220 kilometres per second at the LMC’s distance
of 170 000 light years.

For our Galaxy to keep the LMC orbiting at this speed, it must weigh 10
times as much as all its stars and gas – about 600 billion times the mass of
our Sun. Its halo of dark matter must also extend between 600 000 and 800
000 light years, which means that the LMC is inside the halo for most of
its orbit. Lin and his colleagues conclude that the Magellanic Clouds,
together with the 10 or so other dwarf galaxies around the Galaxy, will
eventually merge with it.

Colin Aspin of the Joint Astronomy Centre in Hawaii and his colleagues
found the protostar in NGC 1333, a nebula 1100 light years from the Earth
which has long been recognised as a region of star formation. In 1983, the
infrared satellite IRAS found seven bright knots in a region to the south
of the optical nebula.

Some of the sources are associated with visible young stars, but others
are embedded deep in the dust clouds. One of these, called IRAS-4, is the
coldest of the sources. This is the object which the astronomers believe
is the youngest star yet found.

Aspin and his colleagues observed the object with two separate telescopes
on Mauna Kea in Hawaii: the United Kingdom Infrared Telescope (UKIRT) and
the James Clerk Maxwell Telescope (JCMT).

The JCMT revealed that IRAS-4 is surrounded by a shell of dust which
is thicker than any yet seen around a young star. This suggests that IRAS-4
is still forming, say the astronomers. Friction is heating the dust as gravity
swirls it down onto the object. It may still be hundreds of thousands of
years before the core becomes hot enough to begin nuclear fusion, say the
astronomers. They believe the protostar is no more than a few thousand years
old.

Some infrared astronomers regard protostars as the Holy Grail. Other
announcements of protostar discoveries since the IRAS survey have turned
out to be fully fledged young stars that were simply obscured by dust clouds.

There seem to be two objects in IRAS-4, both of which are protostars,
say the astronomers. The objects appear slightly elliptical in shape, which
suggests that a disc of dust is orbiting the centre of each protostar and
falling down towards it. Planetary systems could eventually form out of
the remains of these discs, say the astronomers.

According to current theories of star formation, once nuclear fusion
starts, the star will be able to support its outer layers against collapse
through the radiation pressure of the light it emits. Some matter will be
blown off again in the stellar wind. This will eventually clear the dust
clouds away revealing the new star.