Members of the Joint Chiefs of Staff, who were publicly split over President Reagan's last basing plan for the MX, yesterday unanimously endorsed the new proposal to put 100 of the highly accurate missiles in Minuteman silos.

In testimony before the Senate Armed Services Committee, the nation's five highest ranking officers acknowledged that those silos theoretically are vulnerable to Soviet missile attack.

But they are the easiest places to deploy the MX, and the chiefs said early deployment is crucial so the United States can threaten Soviet missile silos and "leadership" in the same way large Soviet missiles threaten this country.

In keeping with this, Joint Chiefs' Chairman Gen. John W. Vessey Jr. and Air Force Chief of Staff Gen. Charles Gabriel pointedly left open the possibility that the United States might not decide to "ride out" any first strike by the Soviet Union.

Sen. J.J. Exon (D-Neb.) said that it sounded as if they were talking of a policy of "launch on warning" in which the United States would fire its missiles upon warning of an attack rather than chance losing them by waiting until Soviet warheads detonated.

Gabriel said that is "not necessarily" the case.

But three times during the hearing Vessey and Gabriel indirectly raised this possibility under questioning.

At one point Gabriel said "They the Soviets don't have any assurance that we are going to sit and ride out an attack." At another he said "if we choose to ride out an attack . . . ." And Vessey, asked by Sen. Edward M. Kennedy (D-Mass.) how many Minuteman missiles might survive a Soviet attack, said, "If we rode out the attack, probably 30 percent would survive today," with levels becoming lower in future years.

The United States has never embraced launch on warning because it raises the possibility that a nuclear war could start by mistake. But it never has ruled out launch on warning, either, to keep the Soviets guessing about what a first strike might cost them.

The question of whether both sides may now be moving toward such as policy has been revived because the 10-warhead MX--especially in vulnerable existing silos--and the new Soviet multiple-warhead missiles are such threatening, yet inviting, targets.

Yesterday's comments by the joint chiefs were among the most explicit hints from high levels that launch on warning has not been ruled out.

And just last week, an Air Force colonel in the MX program embarrassed the administration by reportedly suggesting in Wyoming, just as the president's MX commission was about to make its recommendations, that the United States was moving toward such a policy.

The Pentagon quickly put out a statement saying: "It is our policy not to explain in detail how we would respond to a missile attack . . . . The United States does not rely on its capability for launch on warning or launch under attack to insure the credibility of its deterrent . . . ."

In justification of the MX, Vessey said the latest Soviet missile silos are four times stronger than U.S. silos. This had greatly degraded the ability of Minuteman missiles to destroy those silos, he said.

And Gabriel, asked why it might do just as well to improve the Minuteman as to build the MX, said, "If we don't do MX, no matter what you call it any other missile it would not have the same effect on U.S. allies or the Soviet Union" in terms of demonstrating U.S. resolve to maintain the balance with Moscow.

Asked by Kennedy to make a purely military judgment on the best way to base the MX with all political and technical considerations set aside, the joint chiefs said that was unrealistic and the key thing now was to get the MX deployed and maintain the ability to deter attack.

When pressed, Vessey said that the earlier shell-game plan of the Ford and Carter administrations, a plan that was killed by Reagan, would have added some ability to survive attack.