The Reagan administration's charges that the Soviet Union has used chemical warfare in Southeast Asia have triggered a new debate in the scientific community about the quality of the evidence offered to support the U.S. claims.

Unlike the almost universal acceptance of charges that Iraq used chemical warfare in its war with Iran (it took about three weeks between the charges and generally accepted proof), the Southeast Asian chemical warfare controversy dates back over three years. No definitive resolution is likely soon.

In summary, the United States charged that the Soviets are using a newly created chemical weapon--"yellow rain"--that has killed more than 10,000 people in Laos, Kampuchea (Cambodia) and Afghanistan. The U.S. government has offered evidence ranging from leaf samples to medical and intelligence data in presenting its case.

So far, none of this has satisfied critics of the administration, including some respected members of the scientific community.

The latest difficulty for the State Department's campaign to settle the issue of Soviet chemical war in Southeast Asia is the "bee theory" offered by Harvard biologist Matthew Meselson, an expert on chemical warfare.

Last week, Meselson presented information that he said tended to confirm his belief that the yellow spots of powder associated with the yellow rain attacks have natural causes, and are not part of a chemical weapon.

His statement said his investigations "lead us to conclude that the identification of yellow rain as a chemical warfare agent is a mistake and that yellow rain is actually the feces of Southeast Asian honey bees."

Meselson and Yale University bee expert Thomas D. Seeley recently visited Southeast Asia to collect samples of the bee spots and compare them with samples of yellow spots that the State Department had obtained from the region.

The Meselson-Seeley samples are being analyzed at the Massachusetts Institute of Technology. If they are found to contain the same toxins that showed up in the State Department samples, then a logical conclusion would be that these toxins, or poisonous compounds, are a natural ingredient of bee excrement.

There are eminent scientists on both sides of the issue. But all those outside the government, including supporters of the State Department's claims, say that the department has failed so far to do all the research necessary to settle the controversy.

State Department officials have maintained for years that their evidence is clear and convincing, at least at a common sense level. They say that air-tight scientific proof never may be found, so more research may be useless. In addition, they say that their evidence goes beyond laboratory analysis of leaf samples.

In the case of Iraq, the State Department made its charges about the use of chemical war in the first week of March, and by the fourth week, a U.N. investigation had been conducted and a report written. It found convincing evidence, and condemned those using the chemical agents.

In the yellow rain case, however, the evidence is not as clear cut. Nor is it insulated from the war of words between the United States and the Soviet Union, which has denied any use of chemical warfare.

The issue was first raised to its current pitch by the United States in the fall of 1981, when then-Secretary of State Alexander M. Haig Jr. announced in a speech in Berlin that the United States had found physical evidence of a chemical war campaign in Southeast Asia. He linked the weapons to the Soviet Union.

The State Department case, in brief, said that refugees from the Hmong people and other tribes in Laos and Kampuchea have been reporting the use of chemical war against them for more than five years and the symptoms the chemical causes.

The reports did not match symptoms expected from any known chemical agent. So, sometime in 1980, a government toxicologist devised the theory that the weapon was a new one, a clever use of a natural poison that occurs commonly in the Soviet Union. It is called tricothecene mycotoxin and comes from a fungus that attacks decaying material and stored grains.

State Department officials named four main points of evidence to support the theory: Six or more years of reports from victims and refugees, physical evidence from sites of attacks, medical evidence of a poisonous agent in the blood of victims, and a mass of intelligence data including defectors' reports, radio intercepts, and photos.

Critics say that the refugee interviews are unreliable, the physical evidence is poor and inconclusive, the "victims" could have gotten the toxins by eating moldy food, and that the intelligence reports have not been put out in detail for public scrutiny. For example:

Refugees. The State Department said that several hundred refugees claiming to have been in or near chemical attacks have been interviewed over the past several years. It said the refugee reports have been largely consistent on several key points, including the chief physical symptoms caused by the agent. The problem, critics said, is that only those in one refugee camp, Ban Vinai, and not people out in Lao or Kampuchean villages, have been interviewed. In the camps, they said, stories of chemical war can easily become self-perpetuating legends to which many sicknesses are attributed.

Physical evidence. Virtually all the half-dozen environmental samples from Southeast Asia are yellow or brownish spots comprised largely of pollen. The levels of mycotoxins found in the yellow spots are large, up to about 140 or more parts per million. The State Department and some scientists believe that this variety of fungus and its toxin does not occur naturally in Southeast Asia, so finding it in attack sites provides strong evidence of chemical war.

In addition, there is a Soviet gas mask that the United States got from sources in Kabul, Afghanistan. It had a layer of mycotoxin on the outside, the State Department said. This mask is the only physical evidence of mycotoxin in Afghanistan.

Skeptics said they have numerous criticisms of the physical evidence.

First, the concentrations in the yellow powders are too low to be considered effective enough to be used as a weapon. At 140 parts of mycotoxin for every million parts of yellow powder, it would take a pound or more to cause serious illness or death, according to a conservative estimate by critics. That is far less potent than current chemical agents.

Another point critics make is that no spent shells or other munitions that would be used to deliver the chemical weapons have been found in several years of looking. State Department officials attribute this to the difficulty of obtaining access to the battle areas although enivronmental samples have been gathered from those areas.

On the medical evidence, 60 to 100 people who said they were victims of chemical attack have given blood, urine, or tissue samples. Twenty were found to have mycotoxins in their systems. If the tests are accurate, this proves that the refugees have mycotoxin poisoning, but does not indicate whether the poisoning is from a chemical attack or from moldy food.

Critics said that the samples were collected from sick refugees in Ban Vinai camp in Thailand. But no accurate survey of the other residents of the camp or the food they eat has been done to determine if the cause could be moldy food.

There are explanations that try to reconcile the contradictory evidence.

One suggests that chemical agents have been used, but that they probably are riot gasses used to control crowds. This, combined with some natural fungal poisoning, could provide an answer. On the other hand, if mycotoxins are being used as a weapon, they are not connected to yellow powder but merely stick to the bee spots after an attack and the yellow spots have been picked up as samples of chemical war.

The war of words on yellow rain has not let up. After Haig, Secretary of State George P. Shultz took up the issue with equal determination, declaring that the evidence is overwhelming that the Soviets are engaging in chemical warfare.

Meanwhile, the scientific debate continues.

This is so, many scientists say, because not enough good research has been done to answer basic questions such as whether the toxin involved grows naturally in Southeast Asia and whether the refugees in Thailand are sometimes poisoned naturally by the toxin in their food.

"The preponderance of the evidence is on the government side," said Joseph D. Rosen, a pathologist at Rutgers University who has analysed some of the toxins the State Department says are samples of chemical warfare. "There are still a lot of questions unanswered . . . but it is clear that chemical weapons are being used in Southeast Asia. . . and there is very, very good evidence that the agent, or one of the agents, is mycotoxin."

Political scientist Stuart Schwartzstein, of the private National Strategic Information Center, said, "I have no doubt that chemicals have been used in Southeast Asia" chiefly because of the reports from refugees.

"I don't believe these people are so superstitious that they would just dream up such stories. And no one can kid me that the Hmong people are suddenly being made ill by bee excrement ."

Saul Hormats, a former program director in chemcial warfare at the Army's Edgewood Arsenal in Maryland, disagrees. He said that if the yellow powder is the chemical agent, then it would take about six million pounds of it to attack a village of 12 families. "You would have to be crazy or stupid to make a weapon like that . . . . The Soviets are not that stupid," he said.