ONE DAY IN MARCH OF last year Leonard Tramiel, a balding, dark-bearded man of 45, sat alone in a science classroom in Milpitas, a middle-class community on the south end of San Francisco Bay. Having earned a Ph.D. in physics from Columbia University and having made a bundle in Silicon Valley, Tramiel now taught occasionally around the Bay Area as a volunteer.

Savoring a few quiet moments before 30 eighth graders surged into the classroom, Tramiel opened their astronomy textbook, Prentice Hall's Exploring The Universe, to the lesson for the day. Tramiel was surprised to see that Prentice Hall had inadvertently reversed two photographic images, giving a misleading impression of how the moon looks as it passes through its phases. Tramiel turned back a page. The book said that the moon probably had been born when a giant asteroid had struck the earth, tearing a chunk of material from the planet, and that the Pacific Ocean may be the hole left behind. What was this doing in a science textbook? The asteroid theory hadn't been taken seriously for over 30 years. Tramiel turned back another page and read that the far, or dark, side of the moon had been photographed for the first time by the Lunar Orbiter, a U.S. space probe. He knew for a fact that the Soviets had taken those first photographs.

Three errors in three pages. At home that night, Tramiel read the textbook cover to cover and found dozens of errors--of fact, of interpretation, of concept.

Tramiel called Prentice Hall's 800 number. An operator said someone would get back to him. Nobody did. Calling again three days later, Tramiel insisted on staying on the line until someone helped. A woman identifying herself as one of the editors of the series came on to assure him that the errors had been corrected in a subsequent edition. Tramiel obtained the later edition and found only three of the dozens of errors he had found had been fixed.

Two other Bay Area scientists Tramiel showed the book to, the director of space at NASA's Ames Research Center and an astronomy professor at a local college, termed the Prentice Hall book "horrible."

His interest in textbooks now fully aroused, Tramiel examined a new Prentice Hall book used by his older son's private middle school in Palo Alto. According to the title page, Science Explorer Astronomy had been written by Jay M. Pasachoff, a professor of astronomy at Williams College whom Leonard Tramiel knew to be a renowned scientist. But the book contained such idiocies as references to a "history book from around 800 B.C.," when books did not exist, and to the "rotating"of the earth around the sun, when every schoolboy is supposed to know that the earth rotates on its axis but revolves around the sun. Tramiel checked in with Pasachoff and was not surprised to hear how the errors got in the book--in the editing process, after the book left Pasachoff's hands.

It isn't just Prentice Hall, owned by Pearson Plc., that churns out rubbish for our children to learn by. Scott Foresman, another Pearson company, Holt Rinehart and a range of other publishers are guilty of producing textbooks condemned by experts for their errors and omissions. The whole $4 billion elementary and secondary textbook industry has the problem. In the intensely lobbied textbook selection process in states like California, intellectual content takes a back seat to salesmanship, political correctness, self-esteem for the students and the need to dumb-down lessons so that one product can capture a large market. In the U.S., textbooks sell well if they are designed to hold the attention of children accustomed to MTV and spiffy Internet graphics; they don't sell well if they are challenging. Says Diane Ravitch, a leading historian and analyst of education: "You get a snappy visual package, but it's more like a comic book. The packaging overwhelms the content."

Contrast what goes on abroad. William H. Schmidt, who teaches applied statistics at Michigan State University, in 1995 analyzed 800 math and science textbooks from 50 countries in grades four, eight and twelve, and related the results to learning, as measured by improvements in test scores over the space of a year. U.S. students, and the textbooks they learned from, did not do well at all. The U.S. ranked no higher than 16th on 20 key math areas taught in the eighth grade. In decimals, fractions and percents, U.S. eighth graders ranked 27th, behind their counterparts in such countries as Slovenia, Iran, Latvia, Romania, Iceland and Thailand. Schmidt found that a Japanese eighth-grade math text devotes about 25% of its focus to "congruence and similarity," a core part of geometry. The comparable U.S. book, published by Pearson's Scott Foresman, devotes only 2%. Correspondingly, the Japanese students ranked second in their learning gain in this area; the U.S. ranked 34th.

Center stage in the schoolbook quality debate is California, which centralizes book selection for most of its 8,563 public schools. Given the $500 million a year at stake there, it's no surprise that most of the big publishing houses maintain offices in Los Angeles and the Bay Area, keep lobbyists in Sacramento, and spend millions selling books in the state. Nor is it surprising that Delaine Eastin, California's superintendent of public instruction, has sought and received campaign contributions from textbook publishers. Paramount Communications (now Viacom) even held a glitzy fundraiser for Eastin four years ago on the Paramount lot in Hollywood. At the time Paramount's Simon & Schuster unit had textbook divisions doing millions of dollars of business with the state.

The decision of which texts to buy is in the hands of the18-member Curriculum Commission, made up of school board heads, teachers and other academics around the state. The commission gets advice from two panels of teachers and college professors and its own rulings can be overridden by an 11-member state board of education. Let's look in on the selection process that went on for the 2000-2001 school year science textbooks.

On Mar. 17, 1999 state educational authorities summoned the big publishers to Sacramento for a briefing on the content standards their books would have to meet in order to have a chance for approval. In fourth-grade earth sciences, for example, publishers were told their books must teach students the differences between igneous, sedimentary and metamorphic rocks.

By August the publishers were ready. They submitted 15 sets of science teaching material for California's consideration. The vetting process became contentious. One of the review panels gathered at the Red Lion Inn in Sacramento to vet the submissions of three publishers--Houghton Mifflin; SRA, a McGraw-Hill subsidiary; and Delta Education, owned by Torstar Corp., a Canadian publishing firm.

Delta's submission was a teaching system that relied on gadgets and toys to demonstrate scientific principles. There were no textbooks. In its camp Delta had Kathy DiRanna, a former biology teacher and now an education activist. Opposing Delta: a hard-science contingent of committee members who were disgusted by the early-1990s trend toward fuzzy math and soft teaching and were doing what they could to stop it. The hard-science crowd preferred submissions from Houghton Mifflin and McGraw-Hill's SRA division, for which DiRanna had a barrage of criticisms.

At the same time, another panel heavily criticized Scott Foresman's submission. Little wonder. Scott Foresman diagrammed an electrical current erroneously. It failed to explain the cause of hurricanes. It failed to note that air temperature cannot be measured accurately when the sun is shining directly on the thermometer. It failed to show how to differentiate among igneous, sedimentary and metamorphic rocks.

At that point Scott Foresman, Houghton Mifflin and SRA/McGraw-Hill sprang into action, seeking opportunities to revise their material. Kathryn Costello, the president of Scott Foresman's elementary education group, flew from corporate headquarters near Chicago to meet with Leslie Peterson, president of the Novato school board and an advocate of stricter standards who had led the panel that panned Scott Foresman books.

Costello brought with her Elizabeth Jimenez, a Los Angeles-based Scott Foresman regional vice president. The three women met in the lunchroom of the Novato school district, a low-slung brown building in the small community north of the Golden Gate Bridge. "We want it to work; tell us how it can work," demanded Costello. "What can we do?" asked Elizabeth Jimenez. "We'll do whatever it takes."

Houghton Mifflin, meanwhile, flew in its vice president and national sales manager, Sue Schultz, to try to repair the damage done by Kathy DiRanna. Schultz met with most of the members of the Curriculum Commission, including Richard Schwartz, 56, a high school chemistry teacher from Torrance who has witnessed a lot of watering down of science texts in his 30 years at the school. For example, the current text that he uses, Introductory Chemistry, from Houghton Mifflin's D.C. Heath division, has only four pages on pH, the measure of acidity. A mid-1970s Holt Rinehart text called Foundations of Chemistry had 14 pages on the subject. The Heath text makes made no mention at all of Hess' Law, a critical formula for measuring heat in certain chemical reactions.

The Houghton Mifflin officials tried to assure Schwartz and the other panel members that most of their objections were minor and that Houghton Mifflin could satisfy them. SRA/McGraw-Hill representatives made similar pleas.

While Leslie Peterson was meeting with the two Scott Foresman executives, Leonard Tramiel, the Palo Alto astrophysicist, was 40 miles south in San Mateo, examining more textbooks. He looked through a Holt Rinehart book, Holt Science and Technology, that appeared to be one of the better submissions. Or was it? The book said that protons are helium nuclei; they are in fact hydrogen nuclei. The book said that nuclear energy was first suggested as a possible energy source for the sun in 1899; it was in the 1930s.

Richard Schwartz called the science panel of the Curriculum Commission to order in the early afternoon of Nov. 16, 1999. Leonard Tramiel was the first witness. Without identifying Holt Rinehart, Tramiel said that he had reviewed a text under consideration and found 30 errors in 100 pages. "This morning I checked the [panel's] report on this book. Only 3 of the 30 errors had been found. We cannot trust the textbook industry to correct the mistakes in their books."

The textbook marketers pleaded for time. Prentice Hall, Houghton Mifflin and Scott Foresman officials all offered elaborate plans for mending their products.

But would any real effort be made? During the Christmas holidays a large box from Scott Foresman containing books and teaching materials arrived at Richard Schwartz's home. "Scott Foresman has made an all-out effort to comply with all the changes suggested," Elizabeth Jimenez wrote. "We have also gone further by enlisting the help of Dr. Stan Metzenberg...." The mention of Stan Metzenberg riveted Schwartz's attention. Metzenberg, a biologist at California State University in Northridge, had been a leader, along with Nobel laureate Glenn Seaborg, in the effort to make the standards for California science teaching as rigorous as possible. If Scott Foresman had recruited him, it was a genuine coup.

Schwartz continued reading Jimenez's letter. "We have removed extraneous chapters lessons at all grade levels as identified by Dr. Stan Metzenberg.... All revised new lessons are being reviewed by Dr. Stan Metzenberg...."

Richard Schwartz was impressed. Maybe Scott Foresman had met the challenge after all. He telephoned Metzenberg. "Did you realize Scott Foresman is saying you approved all their programs?"

"What?" an incredulous Metzenberg replied. "I didn't help Scott Foresman. I certainly didn't review their material. I've never seen it."

After speaking to the stunned Metzenberg, Schwartz began perusing the "revised" material Scott Foresman had sent. It was filled with pages that were blank, except for the words "lesson to be inserted." Binders were mangled from hasty packing. Schwartz closed the box like a coroner zipping shut a bagged body.

From his office at Cal State Northridge, Metzenberg dispatched a blistering memorandum to Scott Foresman's Elizabeth Jimenez, with copies to all the officials who had gotten her package, asking that Scott Foresman cease and desist from using his name in promoting its texts. Scott Foresman sheepishly withdrew the science program from California's consideration.

Last January the full Curriculum Commission met to vote on the publishers' submissions. "I call on you once again, as I did in November," implored Leonard Tramiel from the speaker's podium. "Do not recommend these books for adoption. We have no assurance that these materials are accurate."

A majority of the commission voted to reject the SRA McGraw-Hill texts. Approved were several of the other publishers' submissions, including those of Prentice Hall, whose books Tramiel had found so full of errors.

"I have to approve them," Schwartz told Tramiel privately. "The books we have now are so bad, and the new books are significantly better, that I can't in good conscience reject such an improvement."

"If you don't reject 'such an improvement,'Tramiel retorted, "we'll be stuck with them for the next six years until the next science adoption."

On Mar. 16, 2000 the California Board of Education, meeting in Los Angeles and acting on a unanimous recommendation of the Curriculum Commission, voted to approve three kindergarten-through-fifth and sixth-grade science programs, including those produced by Harcourt Brace and Houghton Mifflin, and three programs for sixth to eighth graders, including the programs of Holt Rinehart and Prentice Hall--a total of 6 programs from the 15 originally submitted.

Many of the books, the board acknowledged, still contained errors.