After the Sun has gone down, each lens in a compound eye only receives very
small amounts of light. Most nocturnal insects switch to other senses or have
eyes with large lenses. But bees lack these specialisations. “At night the bee’s
eye is quite hopeless,” says neurobiologist Eric Warrant of Lund University in
Sweden.

To investigate bees’ night vision, Warrant and his colleagues at the
universities of WĂĽrzburg and Constance in Germany used data on the optical
and physiological properties of bees’ eyes to create a computerised simulation
that predicts how well a bee should be able see in any light.

The researchers trained 44 honeybees (Apis mellifera) to associate a
pattern of vertical stripes with a dish of honey. The trained bees were put into
a darkened chamber containing two passages, one with a pattern of horizontal
stripes at its end, the other with a pattern of vertical stripes. They then
reduced the thickness of the stripes until the bees could no longer tell the
patterns apart.

The bees consistently chose vertical stripes that were thinner than the model
predicted would be detectable by eye alone (Proceedings of the Royal Society
B,vol 263, p 1521).

Bees’ brains must use image processing tricks to boost their ability to see
in the dark, Warrant says. There are two possibilities. Just as a photographer
slows the shutter speed of a camera, the brain might lengthen the time over
which photoreceptors collect light before sending a signal. Alternatively,
neighbouring photoreceptors could pool their signals to create a mosaic image
with larger, brighter pieces.

Both strategies have their drawbacks. If photoreceptors collect photons over
too long a period, moving objects begin to blur. And combining signals from
neighbouring receptors improves night vision but reduces image detail.

Warrant suspects that nocturnal bees use a combination of the two strategies,
despite these disadvantages. “It’s a case of being able to see something rather
than nothing at all,” he says.

Normal colour vision is made possible by three types of cone-shaped cells in the retina, which contain one of three pigments. Each pigment responds to either red, green or blue light, although there is some overlap. The brain interprets the colour of an object by comparing the magnitude of the signals coming from the three types of cones. If the three types of cone are stimulated equally then the brain sees grey.

Around 5 per cent of men inherit a form of colour blindness called deuteranomaly, in which the cone that detects green light seems to be absent. But many men with deuteranomaly can still distinguish certain shades of green.

Several theories have been put forward to explain why. Some scientists believe that the stability or efficiency of the pigment might be responsible, or that the nerves which carry signals from the eyes to the brain are faulty.

Jay Neitz, a vision scientist at the Medical College of Wisconsin, Milwaukee, and two colleagues now suggest that the red cones in mildly colour-blind men are sensitive to a greater range of wavelengths than the red cones in men with severe colour blindness. Some of their red cones can detect parts of the green spectrum, thereby compensating to a certain degree for the lack of green cones. The extent of this extra sensitivity determines how colour-blind a person is.

First, the team determined the extent of colour blindness in 16 deuteranomalous men. The men were asked to find a symbol traced out in dots coloured magenta—a mix of red and blue—against a background of grey dots. When a normal eye views the symbol, red and blue pigments are stimulated more than the green pigment, so the brain perceives magenta. But if the eye lacks green pigment, the signals from the other two types of cone are equal, and the brain sees grey.

The men were shown several versions of the symbol, each differing slightly in colour. As expected, the men varied widely in their ability to pick out the various symbols, with some of them compensating perfectly for the lack of green cones and picking out all the symbols.

To explain the range of colour loss, the researchers studied two genes, both of which code for the red-sensing pigment. They focused on specific stretches of the genes which determine the range of wavelengths to which a cone is sensitive by encoding different amino acids in the pigment molecules.

They found that the extent of colour vision loss correlated with differences in the genes coding for red pigment (Science, vol 274, p 801). “For every one of the 16 men, the genes can explain how severe the colour blindness is,” says Neitz.

About 12 000 years ago, a freak cold spell known as the Younger Dryas shut down the North Atlantic Gulf Stream for 1300 years. Without this flow of warm, tropical water to the north, ice sheets advanced again and left much of northern Europe and North America uninhabitable. Climatologists believe that the north could only have warmed up again once the Gulf Stream had restarted.

Last month, at a meeting of the Geological Society of America in Denver, Colorado, Timothy Fisher, a geologist at Indiana University Northwest in Gary, and Derald Smith, a geographer at the University of Calgary, suggested how the Gulf Stream might have restarted. The pair claim that a lake the size of Sweden flooded into the Arctic Ocean. The heat radiating into the atmosphere when the lake water froze in the ocean would have disrupted the circulation of air around the North Pole, “jump-starting” the Gulf Stream.

Fisher mapped the northernwestern extent of Lake Agassiz, which existed in western Canada during glacial times. He found ancient shorelines at different levels left over when the lake surface dropped. Fisher and Smith have calculated that 1.6 cubic kilometres of water drained from the lake every hour for three years along the Mackenzie River, increasing its average winter flow by more than 250 times. “It would have been catastrophic,” says Fisher.

Lake Agassiz had drained several times before, to the east and to the south. But Fisher has provided the first evidence that the lake also drained to the north. The flood started, he says, when the water burst through a natural dam of debris left by a retreating glacier. They dated the draining of the lake to the end of the Younger Dryas by examining wood fragments in flood deposits.

When the lake water reached the ocean and froze, it gave off some 13.5 billion billion joules of energy a day, enough to alter global atmospheric circulation, says Fisher. This would have changed the course of the polar jet stream so that it dipped down over North America before veering back towards the pole, dragging warm water up the coast of eastern North America. Within two to three years, the Gulf Stream current “would have been established and self-perpetuating”.

Other geologists are sceptical about the theory. John Andrews, a geologist at the University of Colorado, Boulder, says that any freshwater added to the Arctic Ocean would have ended up in the North Atlantic and prevented the Gulf Stream from restarting.

Oceanographers used to think that the Irish Sea was constantly flushed by a
current moving northwards at a leisurely 1 kilometre per day. But when Ed Hill,
an oceanographer at the University of Wales in Bangor, tracked drifting buoys by
satellite in 1990, he found a current running 20 times as fast. He was also
surprised to find that it looped back on itself. “It was a chance discovery,” he
says. Hill and his colleagues from the University of Delaware and Britain’s
Ministry of Agriculture, Fisheries and Food presented the findings at a meeting
in Bangor last month.

Most gyres are permanent. They form when tidal or wind-driven currents are
steered by obstacles such as banks or headlands. Hill says that the Irish Sea
gyre is created by summer heating, and only exists between May and October.
“This is the clearest and best evidence yet for seasonal gyres,” agrees Martin
White, an oceanographer at University College Galway in Ireland.

Near the Irish coast, the tides are relatively weak and do not cause much
mixing of surface water with deeper layers. In the middle of the Irish Sea,
however, there is more mixing. So near the coast, the heat of the summer sun
forms a relatively deep layer of warm water. Farther out, mixing with the water
below leaves a thinner warm layer sitting above a dome of cold winter water.

Hill thinks that as the warmer water expands, it raises the sea level by a
few centimetres. Surface water on this raised rim moves by gravity into the
centre, he argues. The rotation of the Earth then curves the path of this water
and sets up an anticlockwise circulation. Hill suspects there may be other
seasonal gyres off the continental shelf of Europe. There are similar domes of
cold water in the North Sea and off the coasts of western Scotland and France,
he says.

The Irish Sea gyre may have some dramatic effects on the region. In the
western Irish Sea, Hill says, the gyre retains the larvae of the Dublin Bay
prawn, Nephrops norvegicus, and the microscopic algae on which they
feed. “If the gyre didn’t exist, the larvae would be flushed out of the Irish
Sea,” he says. Several commercially important fishes, including cod and herring,
also spawn while the gyre is present.

But the current may also be harming marine life by retaining oil and
chemicals spilt into the sea. The gyre may explain the radioactivity recorded at
low levels on the Irish coast north of Dublin, which have been carried there
from the Sellafield nuclear reprocessing plant in England. The gyre should,
however, keep radioactive material from travelling much farther south.

Existing computer models for currents in the Irish Sea need to be revised to
include the surface heating and tidal mixing that form the gyre, says Hill.
“This is crucial if you want to really understand the dispersion of
ł¦´Ç˛ÔłŮ˛ąłľľ±˛Ô˛ą˛ÔłŮ˛ő.”

Until three years ago, many scientists believed that the object must have
been a comet—which would have exploded high above the Earth and done less
damage than an asteroid. But that theory was dealt a serious blow in 1993, when
Christopher Chyba, now at Princeton University, Kevin Zahnle of the NASA Ames
Research Center in California, and others, created computer simulations to
explain the pattern of trees blown down by the explosion. Their calculations of
the explosion’s force suggested that the body involved was a stony asteroid
about 40 metres in diameter, travelling at 15 kilometres per second (New
żěè¶ĚĘÓƵ, Science, 16 January 1993, p 16).

Now Vladimir Svetsov of the Institute for Dynamics of Geospheres in Moscow
has re-examined the American team’s data to estimate the heat of the explosion.
He found that the air around the body would have been heated to 15 000 °C,
enough to melt the surface of the rocky fragments. Droplets would then be
stripped off the fragments until nothing remained. “It was very similar to a
nuclear explosion,” says Svetsov (Nature, vol 383, p 697).

The microscopic debris was probably scattered over a wide area by the
turbulent wake of the asteroid, says Svetlov. The droplets could explain the
origin of spherical particles found in tree resin near the impact site, which
are thought to be extraterrestrial.

Layers of buried animal and plant remains turn into chemicals called
hydrocarbons when they are heated and squeezed. These hydrocarbons are lighter
than the water that fills the pores of the rock, and rise like oil in vinegar.
Over millions of years, oil seeps through pores and fractures in inclined layers
of rock. After journeys of up to hundreds of kilometres, the oil may be trapped
by folded layers.

If companies knew how far oil has travelled to reach known reservoirs, they
would have much better clues about where to explore next. “It really narrows
down the exploration area,” says Alfred Whittaker, a petroleum geologist with
the British Geological Survey in Nottingham.

But the search for a chemical way to estimate the migration of oil has been
difficult. Most of the constituent molecules of oil are unstable and can change
when it is stationary. The heat of the Earth breaks down hydrocarbons and, as
far as 3 kilometres underground, bacteria living in the water-filled pores of
the rock also prey upon them.

Steve Larter, a geochemist at the University of Newcastle, and colleagues at
Shell and several other oil companies, think they have found a group of
molecules that will be more suitable for the job. The molecules, called
non-alkylated benzocarbazoles, are present in oil at concentrations of just a
few parts per million, but they resist most changes as the oil matures.

The fact that the molecule exists in two forms also makes the benzocarbazoles
useful tracers. One form is shaped like a rod, the other spherical. Larter
suspects that the rod-shaped molecules diffuse more easily through the oil and
are more attracted to minerals in the rock. Over distances up to hundreds of
kilometres, more of the rod-shaped molecules will leave the oil.

The team looked at samples of oil from well-studied basins in the North Sea
and western Canada. In both places, the researchers found fewer tracer molecules
in oil that was further from the known source. The proportion also changed: the
shorter the distance the oil had travelled, the higher the proportion of
rod-shaped molecules (Nature, vol 383, p 593).

The technique could help geologists to find new reservoirs in areas where the
source of oil from a reservoir is currently unknown, as, for example, in some
South American basins.

But much remains to be done before the technique is ready for use. For
example, the type of source rock could affect the ratios of the two types of
molecules, say the researchers. And information on the nature of these source
rocks is itself patchy: although distinctive hydrocarbons called biomarkers can
partially reveal the rock’s composition there has been little direct sampling,
because of the cost of drilling.

Another unknown is what happens when oil first leaves the source rock.
Usually, the source rocks are fine-grained sedimentary rocks called shales.
Since these are rich in charted particles of clay, there may be a change in the
ratio of the two molecules when the oil departs, because the rods will be
attracted to the clay.

Despite these remaining questions, Larter says the tracer looks promising.
“We’re proposing this as a possible technology,” say Larter. “It will have to be
tested by the oil companies.”

Archaeologists are keen to know how early humans divided their living spaces
into areas for cooking, storage, rubbish disposal and the like. But fires are
hard to pinpoint because ash often disappears over the ages. “You can still find
the charcoal and flints, but something happens to the ash,” says Ofer Bar-Yosef
of Harvard University.

In the latest issue of the Journal of Archaeological Science (vol
23, p 763), however, Bar-Yosef and his colleagues say that ash does leave behind
a telltale signature that survives hundreds of thousands of years. The
archaeologists’ break came when they teamed up with Steve Weiner of the Weizmann
Institute of Science in Rehovot, Israel.

Weiner analysed sediments from Hayonim Cave in western Galilee, which was
occupied from around 200 000 years ago. Different minerals emit characteristic
spectra when illuminated with infrared radiation. Weiner found that the spectra
of the Hayonim sediments matched those of sediments from a well-preserved hearth
in another cave called Kebara, about 80 kilometres to the north. “That gave me
the idea that they could have been from a fire,” says Weiner.

The researchers suspected that the spectra matched because both samples
contained a component of ash that had not previously been recognised. Further
analysis showed the sediments to contain large amounts of silica-containing
aggregates that resemble siliceous deposits called phytoliths, which are found
in living plants. When the team burnt local trees and shrubs, similar aggregates
made up 2 per cent of the resulting ash.

By studying younger sediments from the caves, the researchers then worked out
how the remaining 98 per cent of ash had changed over time. The initially
dominant mineral, calcite, reacted with phosphates to form a series of soluble
minerals. These were carried away by groundwater, until all that remained were
the insoluble siliceous aggregates.

This process may complicate efforts to date archaeological finds by
thermoluminescence, the researchers say. This technique can determine the time
that has elapsed since a tool was exposed to a flame. The heat of a fire traps
electrons in an object into low energy states, but over thousands of years,
cosmic background radiation excites the electrons back to higher energy states,
which can be measured.

To date a tool, archaeologists correct for the excitation caused by any
radioactive uranium in surrounding sediments. The problem is that when ash
dissolves, radioactive isotopes of potassium become concentrated in the
siliceous aggregates, and this could lead archaeologists to assign dates that
are too recent.

Now the problem is known, it can be worked round. “By detecting these
aggregates, you can tell where you need to be careful,” says Jack Rink, a
geologist at McMaster University in Hamilton, Ontario, who has worked in Hayonim
Cave.

Weiner and Bar-Yosef will now try out their fire divining technique on
sediments from the Zhoukoudian Cave in China. This was the home of Peking Man, a
specimen of Homo erectus that is thought to be one of the first humans
to use fire, some half a million years ago.

FOR seven months, journalist Sara Wheeler travelled, in the company of
researchers, across Antarctica. From McMurdo Station, the main American base,
Wheeler hopped onto cargo planes and helicopters to visit field camps across the
continent. Terra Incognita documents her obsession with Antarctica and
its early explorers.

The strongest aspect of the book is her account of life in the field. Seedy
bars, practical jokes and tedium are all features of days in the frozen
vastness. In one episode Wheeler describes a propane-fuelled toilet—called
a rocket shitter—which explosively backfires. Terra Incognita
reveals a quirky world where residents brew beer in fire extinguishers, write
poems about anything white and decorate their tents with inflated plastic palm
trees.

The book is also packed with historical information. Wheeler appears to have
read every blubber-smeared diary to have come out of Antarctica. In compelling
fashion, she retells the expeditions and sufferings of the great polar
explorers, such as Scott, Amundsen and Shackleton. While her treatment provides
a useful potted history, the jumps she makes between past and present sometimes
prove distracting. The transitions are often awkward and many feel
contrived.

As a visiting writer, Wheeler received support from both the National Science
Foundation in the US and the British Antarctic Survey. Despite constant contact
with researchers, however, she does not seem particularly interested in
Antarctic science. Most of her descriptions of research projects are cursory and
seem to lack a sense of wonder. Perhaps this reflects her attitude towards
science: “Whatever is said about knowledge for its own sake, the only
justification for science in any sphere is that it is a tool of improvement . .
.”

Or perhaps it is simply that for Wheeler, Antarctica’s impact has been more
spiritual. Down there, she says, “I could find the space to look for the higher
power, whatever it was, that loomed over the snowfields.”

Much as I tried, I failed to muster much interest in Wheeler’s spiritual
quest. The anxieties she describes, such as her fear of melancholy and of losing
religious faith, while sincere, are not generally illuminating. Terra
Incognita may reveal an intensely personal relationship with a continent,
but can make the reader feel left out in the cold.

Other 3D techniques, such as holograms and stereoscopic displays, can only be
viewed from a limited range of angles. The headsets used for virtual reality
displays are clumsy, and cathode-ray tubes still provide only a flat surface,
however subtle the colour and shading on the image.

The prototype display uses a pair of diode lasers to trace out images in
three dimensions inside a glass cube. According to the developers, the images
can move in real time and are bright enough to be seen under normal lighting.
The team, which includes researchers from Stanford University, IBM’s Almaden
Research Center and laser company SDL Corporation, reported their success in
Science (vol 273, p 1185).

The concept is not new. In the 1970s researchers at Battelle Laboratories in
Columbus, Ohio, created faint points of light in a crystal of calcium fluoride.
But their display was held back by the limitations of early lasers, which were
bulky and not powerful.

Instead of a crystal, the new system generates images inside a block of
heavy-metal fluoride glass, which is cheaper and easier to produce. The glass
contains traces of rare earth elements, also known as lanthanides, which emit
visible light when stimulated by lasers. The team picked three elements:
praseodymium for red light, erbium for green, and thulium for blue. When
stimulated, they glow for between 0.8 and 15 milliseconds, which is just right
to avoid ghost images or smearing.

Two lasers cast invisible, infrared beams through the glass cube at different
wavelengths. Neither beam has enough energy to make the ions emit light. But
where the beams intersect, their combined energy knocks electrons into a higher
energy level. When they drop back to their ground state, they emit visible
light.

The crossed beams are moved around inside the glass by rotating mirrors
controlled by computer. Although each 3D picture point, or voxel, lights up for
only milliseconds, the lasers retrace the image between 30 and 60 times a
second, which is faster than the eye can detect. The complexity of the images is
limited by how fast the laser can switch on and off to create separate voxels.
The prototype can only manage relatively simple 3D geometric figures.

One difficulty that must still be overcome is that the lanthanide ions must
be isolated from each other in separate sections of glass. When mixed together
they interfere with each other and do not respond properly to the lasers. This
means that the prototype can make red, blue or green images, but it cannot mix
them.

But Elizabeth Downing, a mechanical engineer at Stanford, thinks the team may
have the answer. If they can make the glass block by stacking glass plates a few
hundred micrometres thick, they could have separate red, blue and green voxels
that fool the eye into seeing mixed colours.

Stephen Benton, a physicist at the Massachusetts Institute of Technology who
works on 3D displays, believes the Californian team is on the right track. But
he cautions that they still have a big obstacle to overcome—making the
activated voxels opaque. “The images are wispy and transparent,” he says. That
makes the system unsuitable for photographic images.

But Downing says that the translucent images could be an advantage in some
applications. For example, doctors could look through images of tissue to
underlying organs.