Samuel A. Adams, whose theories about what went wrong in Vietnam are at issue in retired general William C. Westmoreland's $120 million libel suit against CBS, was telling the jury last week about an emotional moment, the day he walked past the thousands of names carved in the dark granite Vietnam memorial.

As he recounted his visit of a year ago, the former Central Intelligence Agency analyst said he tried to determine how many of the dead were killed by the villagers whom Westmoreland had labeled "nonsoldiers" in Vietnam.

"I said to myself, trying to figure the odds, that it was probably at least a third, maybe a half," said Adams, a co-defendant, who was hired by the network as a consultant in preparing the 1982 documentary at issue in this trial.

It was a good pitch for CBS, a moment when a real war was glimpsed through what often has seemed the rehashing of a dry bureaucratic battle. It was one of the best scenes in this long courtroom drama since Westmoreland took the opposite tack two months earlier from the witness stand.

After four days of testimony by Adams -- the first live defense witness for CBS -- it is clear that in many ways the real combatants in this trial are the tenacious former CIA analyst and the stately former general.

Although their views of the CBS program, "The Uncounted Enemy: A Vietnam Deception," appear to be directly opposite, their similarities as witnesses have been striking.

From the witness stand, each has made his pitch as a patriot. Although Westmoreland was more formal and commanding while Adams resembled a folksy scholar, each has spent much time defending what sometimes seemed an outdated concept of honor.

Westmoreland talked about the honor of his troops, how "a commander could have expected no more than they gave." Adams glorified "the sanctity of evidence," summarizing his charges at one point by saying, "We, in intelligence, tried to fool the American public . . . and even to some extent the administration, but we ended up, I think, in fooling ourselves."

Both true believers, each became somewhat disoriented when challenged. Both were strongest as witnesses when their own lawyers pitched them favorable questions.

Part of the change from confidence to confusion on the witness stand is natural, but some witnesses have shown little visible difference in manner whether answering friendly questions or fending off hostile ones.

Adams, 51, and Westmoreland, 70, have presented such strong personalities under their own lawyers' queries that their confusion and apparently conflicting testimony under enemy fire have been all the more noticeable.

Adams, for example, talked passionately about how the home militia, which Westmoreland's command dropped from the official enemy-troop count as "civilians," had caused injuries in Vietnam.

"I remember going into the province hospital in Long An," he said. "A Utah surgeon . . . led me around from bed to bed, and I started asking him about what these things were caused by. He told me that . . . a lot of people had lost legs, lost feet" from booby traps set by civilians. These same troops, who also used homemade mines and punji sticks, played important roles in Vietnam, he said.

During World War II, he said, about 3 percent of U.S. casualties came from mines and booby traps. "In Vietnam, the equivalent percentage was 33 percent -- 11 times higher," Adams told the jury in arguing that Westmoreland should have counted home militia troops among the official enemy.

But under cross-examination, Adams was shown testimony he gave in 1973 on behalf of Daniel Ellsberg in his case involving the Pentagon Papers. Adams acknowledged saying at the time that it was difficult to decide who was military in the Vietnam war. "It was very difficult to decide who to count," Adams testified 12 years ago.

The CBS show charged that Westmoreland was part of a conspiracy in 1967 to suppress higher enemy-troop data to maintain U.S. support for the war. The program accused the general's command of eliminating certain "home-guard" troops from the official roster of enemy troops to keep estimates of enemy soldiers below 300,000. Adams said he and other CIA analysts thought that enemy troop strength at that time exceeded 500,000.

Adams, witty and at ease under questioning by CBS lawyer David Boies, grew more defensive and sometimes irritated and occasionally seemed befuddled Wednesday when facing one of Westmoreland's lawyers, David M. Dorsen. His easy command of numbers seemed to fade as he said more than once that he could not answer without referring to his voluminous notes.

Asked about a memorandum that Adams allegedly has kept for 15 years and that Dorsen describes as supporting Westmoreland about one category of troops, Adams answered: "If it does, I don't recall it. Maybe I have it. Maybe it says so. I don't recall anything like that. I don't think that that was in the memorandum. It may have been."

Dorsen, who is trying to prove that Adams and the other CBS co-defendants believed only those sources for the program who supported their thesis, asked Adams on Thursday why he believed two former intelligence officers whose troop-infiltration estimates were higher than those of other officials.

Adams said he believed that documents showing higher infiltration "had apparently been destroyed," he could not say by whom.

"Did it occur to you that these gentlemen might have been mistaken?" Dorsen asked Thursday, referring to the two intelligence officers.

"I suppose it occurred to me," Adams said, adding, "If I had my doubts in the beginning, they went away" after talking to the two.

Dorsen sometimes brought the trial into an area reminiscent of Alice's Wonderland. Charging that Adams based one crucial figure in the broadcast on a "guess" from his source, Dorsen bore in on the former CIA analyst, who argued that "guess" can mean "best guess" or "estimate" in the intelligence business. Or, Adams said, it can mean "only a guess."

"Is 'only a guess' something less than a guess?" Dorsen asked.

"Only a guess is only a guess," Adams responded, smiling.

Still, Adams' powerful start under questioning by Boies was not demolished by Westmoreland's lawyers. With one more day expected on the stand, Adams, much like Westmoreland, has stumbled but has not fallen.

Even while being cross-examined, Adams managed to make a few solid points. Asked about his part in preparing for a September 1967 session between the CIA and the military, Adams told about complaining to the U.S. Board of National Estimates about what he thought was "falsification" of U.S. intelligence.

A board member reportedly asked Adams: " 'Sam, have we gone beyond the bounds of reasonable dishonesty?'

"And I said, 'Sir, we went beyond the bounds of reasonable dishonesty last August,' " Adams testified.

August 1967, according to Adams, is the month he believes that Westmoreland told his officers that they could not use an official enemy-troop estimate above the 300,000 figure reported by the news media at the time.