"Often, it's when you try and take too much of the pillar" that bumps occur, he said. "I understand they were retreating out of that section and taking some of that pillar out on the way out, maybe shaving it or splitting the pillar to get the coal. The more you take, the greater the load on the rest of the pillar."

And that is what may have led to the Aug. 6 bump, which registered 3.9 on the Richter scale.

"That's my opinion. (Miners were) taking too much" of the pillar, Ferriter said.

"They had mined 80 (feet) by 80 over on the north side, and then they had bumps over there and deteriorating conditions, and ceased mining over there back in March of '07 and moved to the south side. And the consultant recommended that those pillars be increased in size to about 130 by 80.... They recommended it. I don't know that they actually did — and I don't know that anybody actually knows right now."

Prior to Thursday's tragic event, rescue efforts inside the mine had been dangerous and difficult due to the continuing bumps. They created constant problems for rescuers trying to drill through debris to reach the chamber where the six men are believed to be trapped.

The University of Utah Seismographic Stations recorded a 1.6 Richter seismic event at 6:39 p.m. Thursday. It was extremely shallow, U. spokesman Lee Siegel said Friday. There is some uncertainty, but it was believed to have hit about 1/10th of a mile below the surface. The seismic wave forms are consistent with downward motion, just like the original 3.9 collapse. A natural earthquake does not have that signature of downward motion.

"These events seem to be related to the ongoing settling of the rock mass following the main event," Siegel said.

According to the university, 22 seismic "after-events" had been recorded within about two miles of the mine as of Friday morning. Twelve of those were within two days of the original collapse. However, the seismograph stations installed five new seismometers near the mine, so more of the small seismic waves now are being detected. Thursday night's deadly "bump" was detected both by the new seismometers and by part of the university's pre-existing seismic network.

"These events seem to be related to the ongoing settling of the rock mass following the main collapse on Aug. 6," said Relu Burlacu, network manager for the seismograph stations.

Mine owner Bob Murray has disagreed with seismologists over what caused the Aug. 6 mine collapse.

Reviewing a year's worth of seismic events has found less than 2 percent of the seismic events in Utah's coal country are natural earthquakes. The other 98 percent appear to be mining-induced.

"The mountain that the mine's under seems to be settling/collapsing in slow motion," said Siegel.