So far, scientists at the Bower lab, working in collaboration with Gregory Kovacs at Stanford's Center for Integrated Systems, have developed probes with 16 electrodes. But their goal is to produce devices with 1000 electrodes in order to capture the full flavor of the brain in action.

"To record from one thousand neurons is the magic number. It's kinda like let's go to the moon," Kewley laughs. He concedes, however, that the goal may not be practical in the near-term. Listening in on -- and making sense of -- that many conversations at once may be like trying to understand everything being said at once at a big party or political convention.

How much territory would be covered with 1000 electrodes? "Suppose you're recording from Purkinje cells in CrusIIa, a small part of the rat cerebellum that interests us. It deals with whiskers and the upper lip and is about 1 millimeter by 1 1/2 millimeters in size. If we could record one thousand neurons arbitrarily closely spaced we might be able to totally cover CrusIIa. But that's a lot of holes, creating a lot of circuit disruption."

Indeed, the lab is wrestling with the problem of disturbing brain tissue. The brain has billions of neurons crammed inside the skull. Trying to insert these probes, minute as they are, can be like jamming a telephone pole into a bed of flowers. "People are getting to the point where they can record from neurons 300 microns down in tissue [fairly deep in into a rat's cerebellum] but there are problems, big problems What's more, the brain is wrapped in a protective covering --the dura mater (Latin for "hard mother")-- with the consistency of leather. If you try to press something blunt through that covering, you can dimple the surface -- and this dimpling distorts the cells under it. "We're grappling with how to shape the electrodes to avoid dimpling," says Kewley.

Right now, they're only doing these experiments on anesthetized animals. But even when the animal is immobile, inserting the probes can be tricky and the drugs used for the anesthesia do alter brain function. Consequently, they'd like to implant the probe arrays in animals permanently, enabling them to monitor the neuron nets over an extended period of time. "But that raises a whole number of other issues," he adds. "How do you insert it, keep it in place, and make it durable?"

-- Linda Marsa