STOCKHOLM, Sweden (AP) _ Swedish physicists said today they produced a burst of neutron radiation in a room-temperature experiment, repeating the work of two chemists who claimed to have achieved cold fusion in a jar.

The scientists at the Manne Siegbahn Institute for Physics said their method was similar to the experiment that chemists Stanley Pons and Martin Fleischmann said they conducted at the University of Utah.

In March, the two claimed to have produced sustained fusion in a jar at much colder temperatures than previously thought possible. Repeated attempts by other scientists to duplicate the Fleischmann-Pons experiments have failed.

Birger Emmoth, Magnus Jandel, Irena Gudowska and Waclaw Gudowski said in a press release they ran an electrical current between platinum and palladium electrodes immersed in heavy water containing deuterium, a variety of hydrogen.

''The results indicate fusion reactions can occur at low temperatures in the electrode material palladium. This gives certain support to Fleischmann and Pons ideas,'' the release said.

The results were reported in a paper to be published by the scientific journal Physica Scripta. It did not say when their findings might appear in the journal.

In their tests, the scientists obtained increasing neutron radiation reaching a level up to 10 times greater than natural background radiation and lasting up to four hours.

The group said the experiment is difficult to repeat because many factors can affect the process. In some case their tests produced no surplus neutrons.

Birger Emmoth told Stockholm daily, Svenska Dagbladet, the tests were easier to repeat using new palladium plates of sufficient size, rather than already used plates and small ones.

Electrodes used by the group were three millimetres thick and weighed 40 grams, Svenska Dagbladet said.

''Also the character of a time limited 'burst' of course are in line with (Pons and Fleischmanns) results, even though theirs were longer, almost 24 hours,'' he said.

Their experiments, in cooperation with the Royal Institute of Technology and Arrhenius Laboratory in Stockholm, began in early April.