One June morning four years ago I received a telephone call from a lawyer in Greenville, S.C. He was frantic. His client, Albert Thompson, had been convicted of killing a rural store owner and sentenced to death.

No one doubted that Thompson, then 20, had shot the proprietor twice while robbing his store. He and his two accomplices hadn't planned to hurt anyone, but during the robbery Thompson panicked and fired. Thompson and the other robbers were black: The victim was white. Thompson's all-white jury required only 90 minutes to sentence him to death.

In South Carolina (and almost all of the other 36 states with death penalty laws) a murder committed during a robbery is a crime that can land one on death row. That's what the law says, but in fact almost no one actually gets sentenced to death for doing what Albert Thompson did. Out of 195 people who'd been arrested and charged in robbery-murders during the first four-and-a-half years that South Carolina's death-sentencing statute had been in operation, only Thompson and three others -- two blacks and one white, all convicted of killing white victims -- had been sentenced to death and had their death sentences upheld. Four more prisoners' cases were still pending, and all of the remaining 187 had been spared.

Some avoided death because the prosecution allowed them to plead guilty in exchange for a prison sentence. For others, the jury voted for life imprisonment instead of death.

So under the surface of Albert Thompson's seemingly simple case were some gnawing questions. Was he sentenced to death because of his crime, or because of the skin color of the man he killed? Do Southern prosecutors and juries place a higher value on white life than black life? If not, why were the handful of death sentences like Thompson's reserved for the killers of whites?

For his part, Thompson's prosecutor was just doing his job. Before the trial he told Thompson's lawyer that he'd probably be willing to accept a life sentence if Thompson would plead guilty. He just had to check with the victim's family to make sure they agreed.

But the victim's family did not agree, and the prosecutor withdrew the offer. Thompson's lawyer pleaded with one of the victim's relatives. He pointed out that both of Thompson's accomplices were to receive prison sentences, and he asked why the family wouldn't agree to the same for Thompson. Nothing personal, the relative replied, but as far as he was concerned, "all three of the black sons-of-bitches ought to get the chair." That was the end of Thompson's plea bargain. Now his life was in the hands of the jury.

Did Thompson's jury consider race in sentencing him? Probably not. But we all empathize most easily with people like ourselves, and as the jurors looked out over the courtroom, it was the victim's family, not Thompson's, who looked like them.

Moreover, Thompson's jury didn't know about all of the scores of other robbery-murders that had ended in prison sentences. The jurors knew only about this one. They knew that Thompson had shot a man to steal his money, and that seemed enough for them.

That was the process by which AlbertThompson found himself on death row. At each twist and turn race might well have made the difference, not openly, or even as a result of conscious choices: Race plays its part more quietly than that.

If Thompson had been white, or if his victim had been black, maybe the victim's family would have been less intent on a death sentence. Then the prosecutor would have had a freer hand to send Thompson to prison rather than to death row. Or maybe, just for a moment, the jury would have put themselves in Thompson's shoes or in his family's, and shown mercy in sentencing him. But in the coin of the criminal court, Thompson had done something more serious than just killing a man. He had killed a white man.

Albert Thompson won't be executed after all. The state supreme court eventually granted him a new sentencing trial on procedural grounds. By then most people in Greenville had lost interest in his case, and he was quietly resentenced to life. Someone else, chosen for death in much the same way he was, is in Thompson's cell on death row now. The new man has joined 1,800 others -- more than half of them in the South -- who are awaiting execution.

In South Carolina, in Georgia, and in most of the other states of the Deep South, more than 60 percent of all homicide victims every year are black. But with few exceptions, death penalty trials concern the murders of whites, and so do the executions. Of the seven Georgia prisoners executed since the Supreme Court approved Georgia's sentencing system in 1976, all but one have been black, and all but one have been executed for killing a white victim. To know whether a murder poses the risk of punishment by death, the first thing you scan the newspaper accounts for is the victim's race.

Of course, race doesn't affect every capital case. There are some crimes so ghastly that anyone convicted of them is likely to be sentenced to death, no matter what his race or the race of the victim. These are the famous cases, the ones you see in made-for-TV movies or featured on "60 Minutes." Because these are the cases that everyone hears about, they account for the widespread misconception that only the very most heinous crimes ever result in death sentences. It's the "ordinary" cases -- the botched convenience store robberies that make news only in the local paper, the low-visibility cases where prosecutors and juries can easily decide to spare the defendant's life and usually do; those are the cases where race often makes the difference.

Ten years ago, the Supreme Court upheld Georgia's newly-revised death penalty legislation, ruling that the law seemed a colorblind means of picking which prisoners to execute. The Georgia courts have been picking and choosing since then, winnowing the perpetrators of more than 9,000 homicides to just over 100 who now await execution.

This Wednesday morning, having agreed to hear the appeal of one of those prisoners, the Supreme Court will take another look at Georgia's death selection system. The appeal is brought by Warren McCleskey, a young black man sentenced to death for the murder of a white policeman. His argument, based on an exhaustive study of more than 1,000 Georgia homicide cases, is that Georgia sentenced him and many other murderers to death for no better reason than that their victims were white.

Weighting the scales of justice with race is an old tradition in the South. Before the Civil War, Southern statute books graduated criminal punishments according to the race and status of both the offender and his victim. Although the 14th Amendment was enacted to block Southern attempts to recreate this system in the post-Civil War "Black Codes," the actual practice of Southern justice remained largely unchanged for the next 100 years. Southern courts punished blacks with special severity for crimes against whites, but leniency was the rule when blacks committed crimes against other blacks.

In the 1960s, believing that the time was ripe for a challenge, the NAACP Legal Defense Fund investigated death sentences for rape in 11 Southern states. Although the data indicated that race was the principal consideration in determining which rapists were executed, federal courts rejected the argument.

That controversy abated in 1972, when the Supreme Court set aside the death sentences of everyone then on death row. But the court authorized the death penalty for murder again in 1976, and all across the South, the familiar racial patterns re-emerged.

New statistical studies seemed to show that race was again determining who was sentenced to die. But the courts again rejected the studies, asserting that the studies might have overlooked some factor other than race that could have affected the outcome.

Now, however, the most exhaustive study ever performed -- by University of Iowa law professor and statistician David Baldus -- has been placed before the Supreme Court. Taking virtually every conceivable relevant factor into account, Baldus found that race -- above all the race of the victim -- determined which defendants received the death penalty in Georgia. He estimated that a Georgia defendant's chances of receiving a death sentence were more than four times greater if the victim was white than if the victim was black. Baldus concluded that race of the murder victim probably accounted for the death sentences of about a third of the prisoners on Georgia's death row -- including McCleskey's.

The federal appeals court in Atlanta heard arguments based on Baldus' study and rejected them over several vehement dissents. That is the decision that the Supreme Court will review this week.

If this case involved race discrimination in the way Georgia hires prison guards or assigns first graders to public schools, Georgia would almost certainly lose. Ordinary civil rights lawsuits rarely involve evidence of race discrimination as detailed and as powerful as the data on which McCleskey's case is based. The correlation that the Baldus study shows between race and death sentencing in Georgia is two-and-a-half times greater than the proven correlation between cigarette smoking and heart disease. Common sense tells us that an unexplained disparity of that size means that something is seriously wrong.

Warren McCleskey's evidence will also satisfy historians who look back, years from now when the death penalty is no longer in use, for evidence of racial caste bias in Georgia's present death-sentencing system. The Legal Defense Fund's 1965 study of rape death sentences has already made that transition to respectability: Now that the death penalty for rape is just a historical curiosity, no one still bothers to deny that such sentences were routinely imposed according to race. It was only then, when lives hung in the balance, that the courts refused to act on what was already plain to see. The same admission will surely be made some day about death sentencing for murder in Georgia, because by any ordinary standard of proof, McCleskey has proven his case.

But this is not an ordinary case. It's a case about the electric chair, at a moment in our history when the electric chair is popular.

The Supreme Court is supposed to stand as a bulwark against such shifting tides of public opinion. But the court's decisions on capital punishment seem to fit the ebb and flow of the Gallup Poll. In the 1960s and early '70s, when public support for executions was low, the court's decisions cut back on the death penalty, and its 1972 Furman v. Georgia decision cleared every death row in the country on grounds very similar to those urged by McCleskey now. As public support swung back to the executioner in the 1970s and 1980s, the court has tended to step out of the way.

If the Supreme Court holds to that course now, it will rule against McCleskey. The public will conclude that there must not be anything to all these claims of race discrimination, since the Supreme Court didn't accept them. A few months from now, McCleskey's execution will rate two paragraphs on an inside page of your morning newspaper. Violent crime in Georgia will continue to rise and fall as it always has, for reasons having nothing to do with any of this.

And our system of race-weighted death sentencing will have come through yet another challenge intact. The future of that system, 121 years after the end of the Civil War, is the issue in McCleskey v. Kemp, the first civil rights decision of the Rehnquist Court.

David Bruck is a Columbia, S.C. lawyer who specializes in death penalty cases.