A mysterious "flash" signifying a likely nuclear explosion in the far reaches of the South Atlantic, eerily similar to the still-unexplained September 1979 flash in the same general area, was secretly recorded on Dec. 15 by sensitive U.S. monitoring devices.

If the "flash" was indeed a very small nuclear test blast, as is strongly suspected by top intelligence officials who are now certain that that is what occurred in 1979, President Reagan is confronted with one of the gravest, most perplexing mysteries of the nuclear age. The mystery: is the "nuclear club" expanding another notch, or has a card-carrying member taken advantage of the remote waters where the South Atlantic joins the Indian Ocean to test weapons without risk that the world will ever discover who he is?

Whatever the answer, the December "flash" monitored by a U.S. reconnaissance satellite makes one fact clear with deadly logic. With all its monitoring and verification tools, the United States is still unable to solve two enigmatic nuclear riddles, either one of which could affect the course of world history. That surely counsels caution in current comprehensive test ban treaty negotiations with the Soviet Union.

Some Carter administration officials were convinced that President Carter's effort to explain away the 1979 explosion as an event that never happened was a direct result of inability to solve the riddle. To acknowledge publicity that a nuclear test could be conducted even in a remote area without the United States' knowing which country triggered it would undermine Carter's zealous pursuit of the comprehensive test ban treaty. It would also mock his whole non-proliferation program.

So the White House Office of Science and Technology issued its shocking report last July contending that the 1979 event was not a nuclear test at all but rather a chance collision between the reconnaissance satellite and a very small meteor. Specialists in the Carter administration were aghast at this kiss-off of what in fact has created an agonizing dilemma for the United States and a dangerous game for the world: anonymous weapons testing.

The contrary opinion written by the Defense Intelligence Agency was put under a tight "secret" seal after it was read by Carter White House aides. Its still undisclosed findings: the probability is overwhelming that the "event" in September 1979 was no space collision by a low-yield weapons test. Since that DIA report, sent to the White House last spring, new evidence further bolsters that verdict.

As for the second anonymous test Dec. 15, evidence is only slightly less compelling and still being gathered. Indeed, few if any high Reagan officials have yet been informed that U.S. monitoring devices picked up the telltale signals in the same general area -- a vast waterland with a diameter of 3,000 miles -- southwest of the South African coast.

Questions now being urgently addressed in the intelligence community are not confined to the identity of the fugitive state responsible for triggering nuclear tests behind the world's back, although that question is clearly an important one. There is no consensus as to the guilty party, with this exception: it probably is not the Soviet Union.

Intelligence scientists, however, are not unanimous even on this point. It is possible, we were informed, that the Russians lofted the low-yield weapons on balloons from a submarine or trawler and fired them simply to test whether American verification devices were competent to pick up the small blasts.

Most analysts, however, lean to less conspirational theories: that the two tests signaled the entry of a new state into the nuclear club, possibly Israel (known to possess nuclear devices) or South Africa; that they are culminating testing point for nuclear-club member France in developing its own neutron bomb; or that they belonged to India or even Pakistan.

Whoever the villain is that chooses to keep on testing nuclear devices in the atmosphere, the deeper significance to Reagan and his men is this: technology for discovering nuclear testing and verifying performance of states pledged not to engage in certain types of testing is dangerously lagging.

Surely that raises the question of whether the United States should forget about negotiating the comprehensive test ban treaty and at once launch a weapons testing program to catch up to unprecedented Soviet testing the past few years. Although the program appears to be unconnected to the mysterious tests in the faraway South Atlantic, what happened there is a warning signal that the United States may know far less about the difficulties of verification than it is believed. If so, it could be suicidal to rely on good will, good intentions or even signed treaties that cannot be totally verified.