One of the foremost anthropologists of the 19th century was denounced yesterday as a racist who "finagled" his scientific results to depict whites as the superior race.

Samuel George Morton, a Philadelphia physician who ranked the races by the size of their skulls and their cranial capacities, was criticized in this week's Science magazine by Stephen Jay Gould, a paleobiologist at Harvard University. Gould spent the best part of two months reanalyzing the data that led Morton to conclude whites were superior to Indians and blacks. Gould's conclusions are dramatically different.

"There are no differences to speak of among Morton's races," Gould writes in the cover story in Science. "All I discern [in Morton's data] is an supriori conviction so powerful that it directed Morton's tabulations along pre-established lines - his folks on top, slaves on the bottom."

Morton's ranking of the races by brain size has long been forgotten, passed over by the acceptance of Darwin's theory of evolution and permanently supplanted by intelligence (IQ) tests. Gould says he chose to pursue his reanalysis of Morton's work because it raises an issue that scientists often weep under the rug.

"I suspect that truly deliberate fraud is rare in science," Gould writes. "However, I also suspected that unconscious or dimly perceived finagling, doctoring and massaging are rampant, endemic and unavoidable in a profession that awards status and power for clean and unambiguous discovery."

Gould claims there was nothing clean and unambiguous about Morton's discoveries, which began in 1830. By the time of his death in 1851, Morton had collected and analyzed more than 600 skulls from around the world, the remains of eight Caucasian, four Negro and six Indian races.

Morton measured cranial capacity in two ways - by fillings a skull with mustard seed the first time and lead shot the second. His findings were straightforward: Caucasians, especially Teutonic Caucasians, had the largest brains. Indians were next. Blacks were last.

His work was praised as objective. "You work is equally remarkable," Europe's Baron Alexander von Humboldt wrote to Morton in 1844, "for the profundity of its anatomical views, the numerical detail of the relations of organic conformation and the absence of those poetic reveries which are the myths of modern physiology."

Gould describes Morton's work as "suspect from the start."

He says Morton never distinguished male from female skulls even though the female is apt to be smaller because it rests on a smaller frame.

Morton weighted his results by never breaking down skull sizes among Indians and always breaking them down among Teutons and Anglo-Saxons to show their superiority, Gould says. Morton asserted that Teuts and Anglos had the largest skulls, even though his own data showed that Iroquois Indians had larger skulls.

In computing Indian skulls, Morton weighted them down with the skull sizes of Inca Peruvians, whose skulls were small becuase their bodies were small. Gould says Morton deliberately excluded 14 Hindu skulls from the Caucasian sampling for the same reason: their skulls were small and so were their bodies.

Again, he included three female skulls from a Hottentot tribe in Africa in computing average black skulls. Not only were the skulls from the small females, they were from a people whose stature was small than most.

"Morton never recognized the correlation between brain size and body stature, though his own data displayed it so clearly," Gould writes. "Variation among Indian peoples, Hottentots versus taller blacks, Males versus females."

Gould says Morton made other obvious errors. He stated repeatedly that two black skulls were the smallest in his collection even though three Inca Indian skulls were smaller. He excluded a large Chinese skull and a large Eskimo skull from his Mongol collection.

Finally, Morton "rounded out" his average skull size strongly in favor of whites. "He rounded a Negroid Egyptian mean down to 79 cubic inches instead of correctly up to 80 cubic inches," Gould states. "He cited means of 90 cubic inches for Germans and Anglo-Saxons, but the correct values are 88 and 89 cubic inches."

Gould forgives Morton, if only because his fraud was not deliberate. "Morton made no attempt to cover his tracks," Gould says. "He explained everything he did and published all his raw data."