President Carter's top military advisers have decided that the best way to deploy the new MX blockbuster missiles may be to put them on railway cars so they could be moved from place to place, thus making them harder to destroy.

Such deployment is one of two final options described in secret letters from Defense Secretary Harold Brown that were hand-delivered to congressional leaders last night.

The second option, chosen from among five discussed at a White House meeting Thursday, calls for distributing a new hybrid missile among submarines, existing land silos and a new field of holes that would be dug in the West.

Some holes in the new field would contain missiles, and others would not. Soviet gunners, to make sure of destroying every missile, would have to assign one warhead to each hole under this concept, known as the "shell game," long championed by the Air Force.

The hybrid missile would combine the Air Force MX and Navy Trident II. Both are highly accurate and highly destructive.

The basic objective of both options is to offset the threat to existing landbased ICBMs posed by increasingly accurate Soviet warheads.

Carter, though apprised of the options, has not chosen one and is not expected to do so until after meeting with Brown late this month.

Carter and other administration executives have expressed concern about the Air Force "shell game" concept for MX, which originally called for moving 200 missiles among 4,00 holes. The Soviets would have difficulty, critics have said. verifying that all 4,000 did not contain missiles in violation of the SALT II agreement.

Backers of the railroad scheme contend it would solve such verification problems by making it easy for the Soviets to keep track of how many MX missiles were deployed.

The idea is to lay track at the bottom of open ditches on government property. Locomotives would haul cars carrying MX missiles in a horizontal position back and forth between roofed-over stations along the track.

At chosen intervals, the United States could reassure the Soviets by siding back the roof of the stations to show that only a limited number held missiles, thus living up to the missile limits established in the SALT II draft.

The missiles could be moved at night to make it difficult for soviet satellites to see where they went. Another idea is to build covers to slide over the ditches in periods of crisis.

Another advantage of running 200 missiles between 5,000 stations on separate stretches of track, administration officials said, would be the ability to speed the missiles to different stations after soviet missiles were launched.

The network of track for the MX could be located at one government site or in several places, administration officials said. But they estimated it would take an area of some 10,000 square miles to accommodate the 200 missiles and 5,000 stations, built to withstand near misses by Soviet warheads.

Although the Air Force "shell game" would enable the United States to move the lids off the silos at specified times so the Soviets could tell how many held missiles, officials said this poses a much tougher challenge to satellite cameras than the railroad scheme. Also , an MX in a silo could not be moved to another one in a hurry to escape a warhead which was in flight.

Under the SALT II protocol, both the United States and Soviet Union could work on such schemes for mobile missiles through 1981 but could not deploy them until after that. The MX would not be ready before then anyway, according to the Pentagon.

Brown's letters were in the form of a progress report to Chairmen MelvinR. Price (D-ILL.) of the House Armed Services Committee and John C. Stennis (D-Miss.) of the Senate Armed Services Committee. CAPTION: Picture, HAROLD BROWN . . . notifies Hill leaders