One of the researchers involved with that study, Robert Poreda of the University of Rochester, said the newer results need to be interpreted with caution.

"Iridium can be an important tracer for studying impacts, but in some ways the overwhelming signal observed at the K-T site has caused some to view it as a sort of 'key' signal," Poreda said. "At two to three times the background signal [as in Olsen's study] the case can be made for a non-impact scenario."

Nevertheless, Poreda said the new research "represents one of the first steps" in showing that the Triassic-Jurassic extinction was caused by a space rock. His own feeling, he said, is that a few years of study will show that to be the case.

The new iridium study used a special technique called iridium coincidence spectrometry, which finds the element based on the ejection of two gamma-ray particles per atom after they have been irradiated in a nuclear reactor, to make fine comparisons of iridium in the sediment layer compared to older and newer layers.

Christian Koeberl of the University of Vienna performed the work.

Earlier attempts to find an iridium "spike" in the Triassic-Jurassic boundary did not succeed because the spectrometry equipment was not sensitive enough, the researchers said.

The work is painstaking. Earth was an entirely different place 200 million years ago, with all the continents crowded into a single land mass called Pangea. Researchers must first locate exposed sedimentary layers, which may have been folded under and back out to the surface over the eons, and then properly date and sample them.

Much of the evidence was collected in what are now parts of New Jersey and Pennsylvania, along with other dig sites dotting the eastern United States.