science and its short comings

FROM 101 Things You Don't Know about Science and No One Else Does Either BY JAMES TREFIL just three samples

How Much Is a Kilogram?

When you buy hamburger in a supermarket, you aren't likely to worry that the weight written on your package is

incorrect. This is because there is a system stretching from your neighborhood store to scientific laboratories

around the world devoted to making sure that scales are correctly calibrated. Maintaining accurate standards of

measurement has always been a traditional responsibility of governments, and today it is a major scientific

undertaking. But old-fashioned or modern, the basic idea is the samethe government sets the standard for weight or

length or whatever, to which everyone within that government's jurisdiction must adhere.

The oldest such standard we know of is the Babylonian mina, a unit of weight equal to about a pound and a half.

The standards were kept in the form of carved ducks (five mina) and swans (ten mina), and were presumably used

in balances to weigh merchandise. In the Magna Carta, King John agreed that ''there shall be standard measures for

wine, corn, and ale throughout the kingdom." The marshal of the great medieval fairs at Champagne kept an iron

rod and required that all bolts of cloth sold at the fair be as wide as the rod. For most of recorded history each

country has kept various different standards for different purposes. In America, for example, we measure land in

acres, grain production in bushels, and height in feet and inches. According to the Handbook of Chemistry and

Physics, there are no fewer than eighteen different kinds of units called the barrel, for measuring everything from

liquor to petroleum. There is even a barrel used exclusively to measure cranberries!

It was, I suppose, to get away from these sorts of confusions that the nations of the industrialized world signed the

Treaty of the Meter in 1875. According to this treaty, ''the" kilogram and "the" meter were to be kept at the

International Bureau of Weights and Measures near Paris, and secondary standards were to be maintained in other

national capitals. In the United States, they were kept at the Bureau of Standards (now the National Institutes of

Standards and Technology, or NIST) in Washington, D.C. The meter was the distance between two marks on a

length of platinum-iridium alloy, the kilogram the mass of a specific cylinder of the same stuff.

But since the setting of these simple, intuitive standards, advances of technology have made them obsolete. It's all

very well for "the" meter to reside in a vault in Paris, but it would be much more convenient if everybody could

have access to a uniform standard. Thus the trend has been away from the kind of centralized standard-keeping

codified in the Treaty of the Meter and toward standards based on the one truly universal thing we know aboutthe

properties of atoms. The development of the atomic clock is one example of such a move, the new standards for

the meter another. In 1960 the platinum-iridium bar was discarded and the meter redefined as 1,650,763.73

wavelengths of a particular color of light emitted by a krypton atom. Since every krypton atom in the world is the

same, this redefinition meant that every laboratory in the world could maintain its own standard meter. In 1983,

following further development of the atomic clock, the meter was redefined as the distance light travels through the

vacuum in 1/299,792,458 second. Again, this standard can be maintained in any laboratory.

But the kilogram hasn't changed. It's still that same cylinder sitting inside three protective bell jars on a quartz slab

inside a vault in Paris. Even in such an environment, however, atoms of other substances stick to the cylinder's surface. Until 1994 it was cleaned periodically by an old technician using a

chamois cloth. (I remember listening to an absolutely fascinating argument at a NIST lunch over whether or not

removing atoms by washing was worse than letting gases accumulate on the surface.) When the United States

wants to check whether its version of the kilogram still matches the standard in Paris, the American kilogram has

to be carried overseas for tests. The last time this was done, in 1984, two scientists went with it one to carry it, the

other to catch it if it fell.

This is no way to run a high-tech society, and there is an enormous push to develop an atomic mass standard and

put ''the" kilogram into a museum. One technology that may allow us to do this is the new technique of isolating

single atoms in a complex "trap" made of electrical and magnetic forces so that they can be studied for months at a

time. These single atoms stay in the traps so long that they acquire names (the first, a barium atom trapped in

Munich in the 1980s, was called Astrid). It is not too difficult to determine the mass of individual atoms to high

accuracy; the problem is counting the number of atoms in a sample big enough to serve as a mass standard.

The cylinder that now constitutes "the" kilogram contains approximately 10,000,000,000,000,000,000,000,000

atoms, so even if we knew how much each one weighed to incredible accuracy, we'd have a real problem knowing

how many to add. At the moment, at least five different techniques are being developed to give the kilogram an

atomic definition, and I don't imagine it will be long before one of them succeeds. When this happens, the

kilogram will join the meter in its museum.

2] Why Does a Protein Have the Shape It Has?

Life is based on the chemical reactions between molecules, and these reactions depend on molecular geometry.

This statement represents one of the most profound truths we know about the nature of life.

Take the combining of two large molecules as an example. Bonds can be created only between individual atoms.

Think of the atoms that could form bonds as small patches of Velcro on a large, convoluted sphere. For

combination to occur, the molecules have to come together in such a way that the Velcro patches are juxtaposed.

The chance of this happening in a random encounter is pretty small, so chemical reactions in cells depend on a

third kind of molecule called an enzyme. The enzyme attaches separately to each of the molecules taking part in a

reaction, assuring that each one's Velcro is in the right position and allowing the reaction to proceed. The enzyme

is not itself affected in the process. Think of it as a broker who brings together a buyer and a seller but doesn't buy

or sell anything himself.

The enzymes in our cells are proteins, which are long molecules made from a set of smaller molecules called

amino acids. The amino acids are assembled like beads on a string, and the resulting protein then curls up into a

complex shape. Because there are so many possible combinations of ''beads," the final assembly can have many

possible shapes, which makes proteins ideal for the role of enzymes.

Cells work like this: the DNA in the nucleus contains codes specifying the order of the amino acid beads that go

into a protein. The stretch of DNA that contains the blueprint for one protein is called a gene. Each gene codes for one

protein, and each protein acts as an enzyme for one chemical reaction. In a particular cell, as many as a few

thousand genes may be operating at any given time.

This knowledge, combined with our ability to manufacture genes and implant them in bacteria, opens an exciting

possibility. If we know the chemical reaction we wish to drive, we can figure out the shape of the enzyme needed

to drive it. If we know how a specific sequence of amino acids folded up into a protein's final shape, we can design

a gene to make that sequence, put it into some bacteria, and brew it up.

But there's a problem with this notion that has plagued biochemists for the last forty years. Even if we know the

sequence of amino acids in a proteinthe order of the beads on the stringwe simply do not know how to predict the

protein's final shape. A solution to what is known as the ''protein folding problem" remains tantalizingly beyond

the grasp of modern science.

The reason for this gap in our understanding is simple: there can be hundreds of thousands of atoms in a single

protein, and even our best computers aren't good enough to keep track of everything that goes on when the protein

folds up.

At the moment, two lines of research are being pursued. The first involves experiments whose goal is to designate

intermediate states in the folding process. For example, a long chain might first twist into a corkscrew, then have

some segments fold over double, and then fold up into its final shape. By knowing these intermediate states we can

break the folding process down into a series of simpler steps. One difficulty with this approach is that proteins

apparently can follow many different folding sequences to get to a given final state.

Other scientists are trying to use clever computing techniques to predict the final shape a string of amino acids will taketechniques that do not require following each atom

through the folding process. For example, computer programs can estimate the final energy state of different

folding patterns. Since systems in nature move to states of lowest energy, the suggestion is that when you find the

lowest energy state, you have found the final pattern of the molecule. The problem: there may be many low-energy

states, and it becomes difficult to know which one the molecule will wind up in.

Another computer approach involves the techniques of artificial intelligence. Data on known folding patterns of

amino acid strings are fed into a computer, which then guesses a folding pattern for a new molecule based on

analogies to known proteins. The problem: you can never be sure the guess is right.

Whichever technique finally brings us to a solution of the protein folding problem, one thing is clear. When the

problem is solved, we will have eliminated a major roadblock on the road to manufacturing any molecule we want

3] How Does the Brain ''See"?

Take a moment to look around you. What do you see? Perhaps you see a room with colored walls, pictures, doors,

and windows. Whatever you see, though, one thing is cleara collection of cells in your eye and brain has converted

incoming light from the environment into a coherent picture. Over the past few years, scientists have started to

construct an amazingly detailed picture of how that process works.

It begins when light enters the eye and is focused on the retina at the back of the eyeball. There, in the cells called

rods and cones (because of their shapes), the energy of the light is converted into a nerve signal. At this point, the

external environment ceases to play a role in the process, and the mechanisms of the brain and the nervous system

take over. The central question: how are those initial nerve impulses in the retina converted into an image?

Since the early twentieth century, scientists have known that the basic units of the nervous system are cells called

neurons. Each of the many types of neurons has a central cell body, a collection of spikes (called dendrites) that

receive signals from one set of neurons, and a long fiber called an axon through which signals go out to another

set. The neuron is an all-or-nothing, one-way elementwhen (by a process we don't understand) it gets the right mix

of signals from its dendrites, it "fires" and sends a signal out along its axon. The problem for brain scientists is to

understand how a set of cells with these properties can produce images of the outside world.

The first bit of processing of the data carried by the light occurs in two layers of cells in the retina (oddly enough, these cells are located in front of the rods and cones,

blocking incoming light). These cells are connected in such a way that a strong impulse will go to the brain from

one set of cells if it sees a bright dot with a dark surround and a weak signal from that set otherwise. Another set

of cells will send a strong signal if it sees a dark dot with a white surround and a weak signal otherwise. The signal

that goes to the brain, then, consists of impulses running along axons that, in effect, break the visual field down

into a series of bright and dark dots.

These signals arrive at the primary visual cortex at the back of the brain, and from this point on the system is

designed to build the collection of dots back into a coherent picture. The signals feed into a set of cells in one

particular layer of the visual cortex, located in the back of the brain. Each of the cells in that set will fire if it gets

strong signals from a certain subset of cells in the retina. For example, one of the cortex cells may fire if it gets

signals corresponding to a set of dark dots tilted at an angle of 45 degrees. In effect, this cell ''sees" a tilted dark

edge in the visual field. Other cells will fire for bright edges, for edges at different angles, and so on. The output

from these cells, in turn, is passed on to other cells (perhaps in other parts of the brain) for further integration.

So the process of seeing is far from simple. Neurons fire and pass information upward through a chain of cells,

while at the same time feedback signals pass back down the chain and affect lower cells. Working out the details of

this complex chain is one important area of research. Another is following the chain upward, to ever more

specialized neurons. There are, for example, neurons in the temporal lobes (located on the sides of the brain) that

fire only when they get signals corresponding to certain well-defined patternsone set of cells may fire strongly in

response to a dark circle with a horizontal line through it, another to a star shape, another to the outline of a square, and so on. Presumably neurons higher up the

chain combine the output of these cells to construct even more integrated versions of the visual field.

Thus our image of the external world is built up in the brain by successive integrations of visual elements. Where

does the process end? At one point, some neurophysiologists talked about the existence of the ''grandmother

cell"the one cell in your brain that would fire when all the stimuli were right for an image of your grandmother.

While this simple notion has fallen out of favor, it seems likely that within the next decade scientists will be able to

trace the physical connections from the cones and rods all the way to the cells in the brain that put together the

final picture. Other scientists are already trying to locate the cells in the frontal lobe that fire when short-term

memory is activated. This work starts the next step in understanding the visual processfinding out how we

recognize an object whose image has been constructed.

But even when all these neural pathways have been traced out and verified, the most intriguing question of all will

remain: what is the "I" that is seeing these images? To answer that question, we'll have to know about more than

neural circuitswe'll have to understand consciousness itself.