When Paul Alivisatos and his fellow scientists at the University of California at Berkeley succeeded in making invisibly small, atomic-scale cages -- each one a hollow sphere of interlinked cobalt and oxygen atoms with a perfect crystal of platinum rattling around inside -- they basked in the pleasure of knowing they had created an entirely new chemical substance.

The cages -- so small that tens of thousands of them would barely span the diameter of a human hair -- can spur chemical reactions that would not otherwise occur and are the latest in a string of new substances to come out of the hot new field of nanotechnology, which deals with molecules just billionths of a meter in size.

But with their work completed, they faced one more difficult task: what to call their creation.

A team member from China suggested "yolk-shells" because the jiggling platinum crystals resemble wobbly yolks inside eggs. And so it was that the vaguely breakfasty term made it into their scientific report, published in the journal Science last month.

The team's decision was in keeping with a whimsical tradition in nanotechnology, whose first creations -- soccer ball-like spheres made of 60 carbon atoms each -- so resembled the geodesic domes of visionary architect and engineer R. Buckminster Fuller that scientists named them "buckminsterfullerenes," or buckeyballs for short.

But with a steady stream of minuscule new constructs being produced in the world's nanotech labs, including single- and double-walled tubes and kaleidoscopic crystalline spheres resembling disco balls, scientists are realizing that the time has come for a more methodical system for naming their creations.

Regulators at agencies such as the Environmental Protection Agency and the Occupational Safety and Health Administration, who need to know whether certain nanomaterials are toxic and ought to be restricted, also want a systematic taxonomy so they can tell from a molecule's name what it looks like and, ideally, how it behaves.

Even insurers have begun to ask that some taxonomic order be imposed on nanotechnology. A report released last week by Swiss Re, one of the world's largest insurance companies, noted that the daunting task of calculating the potential liabilities of new nanoproducts is made all the more difficult by the lack of a standardized terminology.

Now scientists are tackling the difficult process of creating one. The effort is young; experts are just now organizing a series of conferences to hammer out a system. But the process offers an unusual peek into the arcane world in which chemists decide how to categorize the tangled skeins of new knowledge.

Done right, experts said, a common vocabulary will foster more accurate communication among scientists and others in a field that -- thanks to best-selling thrillers such as Michael Crichton's "Prey" -- is already widely misunderstood. But other fields of science have failed in similar efforts and suffer today in Babel-like confusion -- an indication of how difficult it can be.

"It's like developing a new language, and I don't want this to become Esperanto," said Vicki Colvin, director of the Center for Biological and Environmental Nanotechnology at Rice University in Houston and a prime organizer of the new nano nomenclature effort.

All chemical substances have a simple "chemical formula" -- such as water's famed "H{-2}O" -- listing their constituent atoms. But formulas say nothing about where the atoms attach to one another -- that is, the three-dimensional structure of the molecule. To do that, scientists use terms, including suffixes, prefixes and numbers, that together allow anyone familiar with the language to draw a 3-D picture of the molecule.

By one convention, for example, cholesterol has a chemical name of "delta5-Cholesten-3beta-ol." Alivisatos of Berkeley gave his yolk-shells a similarly complicated chemical name, but the notation he used is but one version of many emerging in the nano field.

The naming of nanoproducts poses novel challenges. Because they are so small, they behave very differently than macro-size substances of similar chemical composition. Some materials that do not normally conduct electricity, for example, do so very well when made small enough. And some materials that reflect light are completely transparent on the nanoscale. So existing names for those materials can be deceptive.

"The problem is, how do you go about naming materials that are chemically and atomically identical to larger structures but clearly have a different activity level?" asked Kristen Kulinowski of the Rice nano center.

Of particular interest to regulators and toxicologists is emerging evidence that some substances that are normally biologically inert can cause worrisome reactions in the body when present as nanoparticles. Similarly, some substances that are normally safe in the environment seem to have the potential to be ecologically disruptive when dispersed as nano-size particles.

So while scientists generally expect that the vast majority of nanomaterials will prove safe, they also want a naming system that does the best job of predicting important traits, including biological effects. If the atomic coatings that scientists often attach to their nanomaterials are key to the material's effects on living cells, for example, then precise terms relating to those coatings will have to be adopted in the same way that Eskimos have many words for different kinds of snow. If a particle's size or the amount of surface area is key to its behavior, then terminology will have to focus on that trait.

In some cases, existing terms may suffice. In others, words may have to be invented, experts said.

Nanomaterials also pose a problem for classifiers because it can be difficult to nail down their precise structures -- a prerequisite to naming them. The problem is that the act of measuring or analyzing these substances can move a few atoms around, which in the case of nanoparticles can be a substantial fraction of what was there.

Colvin hopes to receive funding to begin a series of nano nomenclature meetings this August and expects it could take as long as two years to get a solid framework. These meetings will include biologists, environmental scientists and others, reflecting the many potential applications for nanoproducts.

In the end, experts said, nanotechnology will be a more precise science, if a little less fun -- a science in which, at least according to one proposed system, some old-fashioned buckeyballs may find themselves renamed (C60-Ih)[5,6]fullerenes.

If that isn't progress, then what is?

Titanium dioxide nanomolecules are used in sunscreen. Discussion of standardized terminology is just beginning.