The Pentagon has just begun to test whether it can harden missile silos as much as it says its new "Dense Pack" basing plan for the MX will require.

Defense Secretary Caspar W. Weinberger's own hand-picked MX expert has warned it is "uncertain" such hardening can be achieved, and has also said the Soviets may be able to counter the MX "almost as soon as it is fully deployed."

Whether the Pentagon can harden the silos enough to enable the MXs to survive Soviet attack is crucial in Congress' debate over President Reagan's Dense Pack plan.

That level of hardness is "up to 10 times anything ever before achieved," according to one source involved in the program.

In a Sept. 22 letter to Weinberger classified "secret," Dr. Charles H. Townes, who headed two panels that studied MX deployment, said that his panel concluded even levels of hardness that would resist up to 10,000 pounds of pressure per square inch were "not adequate" for Dense Pack. The latest hardening of U.S. Minuteman II missile silos, by comparison, is intended to withstand about 2,000 pounds per square inch.

The hardness required for Dense Pack, Townes wrote, "has been increased by a large factor" over 10,000 pounds per square inch and its "achievement is uncertain."

Under Dense Pack, 100 giant new MX missiles would be bunched together within superhardened silos in a strip a mile wide and 14 miles long in Wyoming.

The Soviets would be unable to destroy all 100, the administration argues, because the blast from the first incoming Russian missiles would destroy those that followed. This phenomenon, called "fratricide," would permit more than half the MX missiles to survive, according to the plan.

"The most serious problem" with this configuration, Townes wrote, "is that the Soviets may have appropriately modified their weapons for an effective attack on it almost as soon as it is fully deployed." The general thrust of Townes' letter was reported last month, but not his language indicating the depth of his concerns.

Because of the strength of Townes' doubts about Dense Pack, the Air Force sought unsuccessfully to keep his letter out of the package of documents that went to the president before Reagan decided on that basing plan for the MX. Members of Congress recently asked the Defense Department for copies of the Townes letter as part of their effort to determine whether the silos can be hardened sufficiently to permit the Dense Pack basing to work.

A second Weinberger expert, Dr. William Hall of the University of Illinois, disagreed with Townes' view in an interview yesterday. Hall, a structural engineering specialist who has studied earthquake phenomena, said the "requisite hardness can be achieved" but he conceded it would be "pushing the technology . . ., something often done in weapons programs."

Hall said there is "a tremendous amount of testing to be done" on the system, not just for hardness of the silos but also "for the shock isolation system," which would permit the MXs to withstand the earth tremors that would result from the expected explosions.

On Wednesday, the Defense Department released the results of a Nov. 8 test in which a concrete-and-steel scale model of one silo design was subjected to a high-explosive simulation test that was equivalent, in blast effect, to a 25-megaton bomb. The Pentagon press release described the results as surpassing "Air Force expectations in terms of resistance to blast effects, with blast resistance being more than 10 times that of previous United States silo design."

The test, however, simulated a bomb exploding in the air above the missile silo, according to a Pentagon official. Its aim, sources said, was limited to seeing the effect an air burst would have on the silo below it. Thus, it did not simulate the basic hardness problem the MX silos must face from a ground burst of a 25-megaton weapon.

Dr. Eugene Sevin of the Defense Nuclear Agency, which is conducting the tests, was quoted in the Pentagon release as saying the prototype silo used was "a first cut" and that "as a result, this test does not represent the limits of what we might actually do with specific designs."

A second hardness test was conducted Wednesday, according to a Defense Department spokesman. But he would not describe the type of test and would only say it took place "in Kentucky." No more tests are scheduled this year, he added.

The Nov. 8 test "can't be considered a very conclusive test," a civilian member of the Defense Science Board said in an interview yesterday. The prime test, he said, "would come from tests of ground movement" that would be experienced in simulated ground burst tests.

According to this specialist, who supports the Dense Pack plan, the Air Force "has a chance of achieving" the hardness level it is seeking; but its experts "don't have a design yet," and achieving one "will be costly."

Townes, in his letter, raised the prospect of "delays in full deployment" of the MX missile brought on by problems in achieving hardness levels required and by "cost overruns." Thus, he said, it may be "particularly difficult" to deploy the first MX missiles in the Dense Pack plan by the 1986 target date.

Under his analysis, Townes said, the Soviets' costs and technical difficulties in designing and deploying a nuclear warhead to defeat Dense Pack would be less than those facing the Pentagon in designing and building the basing system.

He called the Air Force "scenario" -- in which the Russians will not be able to come up with a new warhead to defeat the Dense Pack system until 1990 or 1991 -- "at best quite uncertain and maybe unlikely."