Most of this work was to lead nowhere. They knew that, and accepted it in advance. As Stone was fond of saying, scientific research was much like prospecting: you went out and you hunted, armed with your maps and your instruments, but in the end your preparations did not matter, or even your intuition. You needed your luck, and whatever benefits accrued to the diligent, through sheer, grinding hard work.

Burton stood in the room that housed the spectrometer along with several other pieces of equipment for radioactivity assays, ratio-density photometry, thermocoupling analysis, and preparation for X-ray crystallography.

The spectrometer employed in Level V was the standard Whittington model K-5. Essentially it consisted of a vaporizer, a prism, and a recording screen. The material to be tested was set in the vaporizer and burned. The light from its burning then passed through the prism, where it was broken down to a spectrum that was projected onto a recording screen. Since different elements gave off different wavelengths of light as they burned, it was possible to analyze the chemical makeup of a substance by analyzing the spectrum of light produced.

In theory it was simple, but in practice the reading of spectrometrograms was complex and difficult. No one in this Wildfire laboratory was trained to do it well. Thus results were fed directly into a computer, which performed the analysis. Because of the sensitivity of the computer, rough percentage compositions could also be determined.

Burton placed the first chip, from the black rock, onto the vaporizer and pressed the button. There was a single bright burst of intensely hot light; he turned away, avoiding the brightness, and then put the second chip onto the lamp. Already, he knew, the computer was analyzing the light from the first chip.

He repeated the process with the green fleck, and then checked the time. The computer was now scanning the self-developing photographic plates, which were ready for viewing in seconds. But the scan itself would take two hours—the electric eye was very slow.

Once the scan was completed, the computer would analyze results and print the data within five seconds.

The wall clock told him it was now 1500 hours—three in the afternoon. He suddenly realized he was tired. He punched in instructions to the computer to wake him when analysis was finished. Then he went off to bed.

In another room, Leavitt was carefully feeding similar chips into a different machine, an amino-acid analyzer. As he did so, he smiled slightly to himself, for he could remember how it had been in the old days, before AA analysis was automatic.

In the early fifties, the analysis of amino acids in a protein might take weeks, or even months. Sometimes it took years. Now it took hours—or at the very most, a day—and it was fully automatic.

Amino acids were the building blocks of proteins. There were twenty-four known amino acids, each composed of a half-dozen molecules of carbon, hydrogen, oxygen, and nitrogen. Proteins were made by stringing these amino acids together in a line, like a freight train. The order of stringing determined the nature of the protein—whether it was insulin, hemoglobin, or growth hormone. All proteins were composed of the same freight cars, the same units. Some proteins had more of one kind of car than another, or in a different order. But that was the only difference. The same amino acids, the same freight cars, existed in human proteins and flea proteins.