Nuclear war is generally believed to bring risks of destruction out of proportion to any gain that may be secured by the war, or to any loss that may be averted, except perhaps for the loss of national independence and group survival. Nuclear-armed states, however, continue to project military force outside their own territory in order to carry out rivalries for power and influence. Will these rival power projections lead to war, as they often did in the past? If not, how will they be resolved? This paper makes the case that, because of the recognized destructiveness of nuclear weapons, rivalries among major nuclear-armed states for power and influence outside their own territory are not likely to lead to central war among them, but that definite lines separating zones of exclusive security influence, such as prevailed during the Cold War, will reappear where circumstances prevent
other compromises. This conclusion does not hold in the case of nuclear powers that are centrally vulnerable to conventional attack from each other: in that case, nuclear deterrence is less likely to be stable. Where lines are established, they may facilitate rather than prevent cooperation in dealing with the next century's global problems.

It is no exaggeration to say that arms control has undergone a revolution in the past decade. In the forty years since the bombings of Hiroshima and Nagasaki ended World War II and began the nuclear age, governments, organizations, and individuals have worked to reduce the threat of wars between great powers employing weapons of mass destruction-nuclear, chemical, and biological. Some progress was made during this period; the Limited Test Ban Treaty (LTBT) of 1963, the nuclear Non-Proliferation Treaty (NPT) of 1970, and the Anti-Ballistic Missile (ABM) Treaty and Biological Weapons Convention (BWC) of 1972 were the more notable achievements. But progress was always slow, frustrating, and tentative, with no assurance that the whole fabric might not be undone by an increase in superpower tension or by domestic forces in either the United States or USSR hostile to the very concept of arms control.

This report summarizes analytical work completed on the Trident SLBM nuclear weapons safety issue. First, we evaluated the increase in low levels of risk of death from cancer from potential plutonium dispersal accidents at the Trident base at Kings Bay, Georgia. Specifically, we estimated the number of latent cancer fatalities resulting from a hypothetical worst-case accident involving a 10-kilogram release of weapons-grade plutonium aerosol at Kings Bay with the wind direction toward downtown Jacksonville, located 55 kilometers away. The estimated number of long-term cancer deaths ranges from 5 to 3300 and depends on a number of factors and assumptions including deposition velocity, wind speed and direction, the nature of the plume, and mixing layer height.

Second, we applied a simple, "back ofthe envelope" risk-analytic approach to the Trident safety problem to try to shed some light on the key question: How much should be spent on safety modifications for Trident? Depending on a variety of assumptions and value judgments, our analysis suggests that if one believes that the probability of a serious accident over the 30-year Trident program lifetime is of order 0.01 to 0.10, then an expenditure of $1-5 billion to increase safety is not unwarranted given reasonable estimates of the consequences of such an accident.

Using major new documentary sources, this book tells the story of why and how China built its nuclear submarine flotilla and the impact of that development on the nation's politics, technology, industry and strategy.

Based on interviews with participants and research in newly opened archives, the book reveals how the American atomic monopoly affected Stalin's foreign policy, the role of espionage in the evolution of the Soviet bomb, and the relationship between Soviet nuclear scientists and the country's political leaders.

The end of the Cold War creates both new challenges and new opportunities for improving nuclear weapons safety. Several post Cold War developments are likely to have negative effects on the safety of existing nuclear weapons arsenals. These potentially dangerous trends include an apparent decline of morale in the laboratories and military organizations responsible for weapons safety, the potential proliferation of nuclear weapons to new states, the likely discontinuation of nuclear testing for safety related purposes, and the introduction of new nuclear weapons operations, including large-scale warhead dismantlement and the relocation and long-term storage of large numbers of nuclear weapons.

In an effort to explore such challenges and opportunities, a NATO Advanced Research Workshop was held in Oxford, England from August 25^th through 27^th, 1994. The workshop produced seven specific proposals for consideration to increase nuclear weapons safety and security. The proposals represent a summary of the points discussed at the workshop and are not intended to imply complete consensus of all participants.