PASADENA, CALIF., OCT. 2 -- A preliminary investigation of Thursday's southern California earthquake indicates that it originated in an underground fault extension previously unknown to seismologists, a discovery that raises questions about traditional methods of detecting dangerous earthquake zones.

The quake, classified as moderate at 6.1 on the Richter scale, killed six people and injured more than 100. It was the strongest earthquake in Los Angeles County in 16 years and the third largest in the area since record keeping began in 1932. Officials today assessed property damage at more than $40 million.

Seismologists with the U.S. Geological Survey and the California Institute of Technology have concluded that the first, jarring shock occurred eight miles underground and more than one mile northwest of the mapped termination of the Whittier Fault, 12 miles east of downtown Los Angeles.

Several of the more than 12 aftershocks occurred as much as three miles north of the mapped fault line, and at a depth that suggests a previously unknown, almost horizontal fault difficult to detect.

The existence of such hidden faults, which scientists said may be fairly common, "becomes a concern in terms of, for instance, siting nuclear power plants," said Lucile M. Jones, a USGS seismologist here.

The newly discovered fault dips at a 25 degree angle beneath El Monte, Rosemead and Temple City, between Pasadena and Los Angeles. A similar underground fault was discovered four years ago in the San Joaquin Valley after a 6.5 magnitude earthquake devastated the downtown streets of Coalinga, Calif., where no serious quakes had been forecast.

Using oil-industry techniques, scientists have begun to look for other hidden faults but acknowledge that their great depth and the expensive equipment required limit their efforts.

Jerry Eaton, a research seismologist at the USGS office in Menlo Park, noted that scientists took a while to accept the notion that the visible areas of slippage, or faults, on the earth's surface signaled earthquake zones. Once that was proven, "we got the notion that if we took our map and put down all the faults we could see, we had done our job," Eaton said.

Now, much more research is necessary, particularly at the frayed ends of visible fault lines where odd, hard-to-detect earth fractures often occur. Jones said the initial data indicates that the Whittier fault, of the common vertical type with visible slippage lines on the surface, plunges underground and twists until it forms a broad semi-horizontal fracture zone eight miles down.

Jones and Eaton said anticlines, hills formed by buckling of the earth's surface, provide one clue to the existence of horizontal faults underground. Caltech scientists have found a graduate student's report on file that reveal anticlines in the area where Thursday's earthquake occurred.

Scientists from Cornell University and other laboratories are attempting to detect such faults and other subtle features of the earth's crust, using heavy equipment that sends vibrations into the earth and delicate instruments which detect the echoes off faults underground.

Such work is slow, and with very little funding available, seismologists said that most such hidden faults are not likely to be discovered until they produce earthquakes such as Thursday's.

"The existence of these faults is no real cause for panic," Eaton said, but it underlines the need for increasing the number of seismograph stations and carefully studying all newly recorded quakes. Haresh C. Shah, a structural-engineering professor at Stanford University, said many surface faults are also hard to detect in the intricate crinkling of the earth along the two continental plates that move past each other on the California coast.

Scientists here said Thursday's quake resulted from a "thrust" movement, in which the upper portion of the horizontal fault zone slipped up slightly. The Coalingua quake was caused by a similar thrust.

A pink dot on a map at the USGS office indicated the quake's estimated epicenter below the corner of Fern Street and Loma Avenue in South El Monte.

Unreinforced brick buidlings on that corner suffered the most damage, and ramshackle wooden houses in the area seemed undisturbed by the quake. Workers at the D, B & A Stanley Pest Control and Don Coniglio Co. at that corner said the shaking at 7:42 a.m. Thursday had been severe, but no serious damage or injury had resulted.

Few Los Angeles area households escaped the earthquake without some minor damage, fallen glasses or light fixtures or small cracks in walls and ceiling.