SEVERAL CENTURIES AGO, some "very light-skinned" people were shipwrecked on a tropical island. After "many years under the tropical sun," this light-skinned population became "dark-skinned," says Biology: The Study of Life, a high-school textbook published in 1998 by Prentice Hall, an imprint of Pearson Education.

"Downright bizarre," says Nina Jablonski, who holds the Irvine chair of anthropology at the California Academy of Sciences. Jablonski, an expert in the evolution of skin color, says it takes at least 15,000 years for skin color to evolve from black to white or vice versa. That sure is "many years." The suggestion that skin color can change in a few generations has no basis in science.

Pearson Education spokesperson Wendy Spiegel admits the error in describing the evolution of skin color, but says the teacher's manual explains the phenomenon correctly. Just why teachers are given accurate information while students are misled remains unclear.

But then there's lots that's puzzling about the science textbooks used in American classrooms. A sloppy way with facts, a preference for the politically correct over the scientifically sound, and sheer faddism characterize their content. It's as if their authors had decided above all not to expose students to the intellectual rigor that is the lifeblood of science.

Thus, a chapter on climate in a fifth-grade science textbook in the Discovery Works series, published by Houghton Mifflin (2000), opens with a Native American explanation for the changing seasons: "Crow moon is the name given to spring because that is when the crows return. April is the month of Sprouting Grass Moon." Students meander through three pages of Algonquin lore before they learn that climate is affected by the rotation and tilt of Earth--not by the return of the crows.

Houghton Mifflin spokesman Collin Earnst says such tales are included in order to "connect science to culture." He might more precisely have said to connect science to certain preferred, non-Western, or primitive cultures. Were a connection drawn to, say, a Bible story, the outcry would be heard around the world.

Affirmative action for women and minorities is similarly pervasive in science textbooks, to absurd effect. Al Roker, the affable black NBC weatherman, is hailed as a great scientist in one book in the Discovery Works series. It is common to find Marie Curie given a picture and half a page of text, but her husband, Pierre, who shared a Nobel Prize with her, relegated to the role of supportive spouse. In the same series, Thomas Edison, inventor of the light bulb, is shown next to black scientist Lewis Latimer, who improved the light bulb by adding a carbon filament. Edison's picture is smaller.

Jews have been awarded 22 percent of all Nobel Prizes in science, but readers of Houghton Mifflin's fifth-grade textbooks won't get wind of that. Navajo physicist Fred Begay, however, merits half a page for his study of Navajo medicine. Albert Einstein isn't mentioned. Biologist Clifton Poodry has made no noteworthy scientific discoveries, but he was born on the Tonawanda Seneca Indian reservation, so his picture is shown in Glenco/McGraw-Hill's Life Science (2002), a middle-school biology textbook. The head of the Human Genome Project, Francis Collins, and Nobel Laureates James Watson, Maurice H.F. Wilkins, and Francis Crick aren't named.