The news dispatch said. "Thieves stole gold, silver and platinum worth an estimated $1.2 million in a weekend burglary of a Paris firm."

When we printed the story, we put a headline over it that said, "Precious Metals Robbed." Chlorinda V. Russo thereupon asked, "How does one rob a precious metal?"

A similar headline appeared in Sunday's paper. Ove a story about bandits who used a helicopter to rob a Montreal bank of $12,000, we put the headline, helicopter Bandits Rob $12,000."

The metals were not robbed. They were stolen. The $12,000 was not robbed. The bank was robbed.

Rob is frequently misused these days. The word refers to the act of taking something by force or the threat of force, as in the case of a robber who uses or diaplays a weapon. Steal , on the other hand, carries the connotation of stealth. There is a fundamental difference between robbery and burglary, and we should not have used rob in a headline over a story about a burglary .

It goes without saying that space limitations make it extremely difficult to devise headlines that say precisely what they ought to say. However, in these two instances we were just plain wrong. There was plenty of space to change the first headline to "Precious Metals Stolen" and the second one to a suitable alternative.

Another headline in last Sunday's paper touched off an internal discussion before it was set into type. Our man in Hong Kong, Jay Mathews, had reported that Rep. Elizabeth Holtzman (D-N.Y.) and Rep. Billy Lee Evans (D-Ga.) had obtained a brief look at the Sino-Vietnamese border war. The headline over that report said, "Two Congressmen Report Battle Signs at Langson, Laocai."

Several copy editors raised their eyebrows. Rep. Holtzman may be a congresswoman or even a (shudder!) congressperson, but she is obviously not a congressman.

Representatives would have been a properly sexless alternative to congressmen , but representatives would not have produced the instant recognition that headline words should have. To Americans, a member of the lower house of Congress is a congressman, not a representative. And we never refer to members of the upper house of Congress as congressmen. They are always called senators.

Headline writers must convey information rapidly, and are therefore as concerned with being effective as with being technically correct. so the gentlewoman from New York remained a "congressman" in our headline.

Dorothy Webster Clark referred me to a recent item in our newspaper that spoke of attackng people with a fork and as a consequence leaving them with four tiny holes in their throats -- as if they had been the victims of "Siamese vampires." Mrs. Clark wrote:

"Not all Siamese are double anything. Only Siamese twins. It is meaningless if the word twin is not also used.

"Newspaper stories show double carrots or tomatoes and call them 'Siamese' carrots or tomatoes. It just doesn't make sense.Siamese doesn't mean double."

A Feb. 18 article about Sharon Pratt Dixon said, in part, "Articulate and poised, she handles difficult questions easily; notheing seems to phase her." A Fairfax reader circled phase and wrote in the margin, "I trust this careless writing and editing is only a passing faze."

Owen J. Remington criticized The Washington Post's reference to the 100th anniversary of Edison's "discovery" of the light bulb. To discover, he points out, is to find or uncover an existing thing. Edison didn't find an existing light bulb. He figured out how to make one. Or he invented it.

Several readers sent me clippings of a Feb. 16 story about industrial production. It began by saying, "The nation's industrial output slowed almost to a halt last month," and the headline above it echoed, "Output at a Virtual Halt." Did that mean what it said? Fortunately, it did not.

The story went on to say that industrial production was 0.1 percent higher in January than it had been in December, and that it had been a hefty 0.7 percent higher in December than it had been in November "after similar increases in each of several previous months."

Then the story went right back to talking about the "falloff in industrial output" and "the slowdown in industrial production." Never once idd it say what it should have said, that although the rate of increase was slowing down, total output remained at an all-time high.