Twilight of the Gods closes out Toll's trilogy on the Pacific War with brutal 18 months of the war. With the comprehensive defeat of Japanese airpower at the Battle of the Phillipine Sea and invasion of Saipan, the war entered a terrible stage of attrition. Japan's defeat was inevitable, but the confused decision making apparatus in Tokyo was incapable of recognizing the fact, and that Roosevelt's 'unconditional surrender' terms were the best deal obtainable. Instead, Imperial Japan embarked on a strategy of attrition, of gyokusai (literally "shattered jewels") attacks with kamikaze suicide craft, and intricate systems of underground defenses. Perhaps the defenders could exact such an intolerable cost that America would sue for peace, rather than send men to die on more forsaken Pacific islands. Perhaps another miracle would step in to save Japan, as it had in history.

To get to Tokyo, the Allied military had to seize three final islands. Recapturing the Philippines was an obsession for General MacArthur, but Admirals Nimitz and King agreed that Leyte and Luzon offered more favorable circumstance than Formosa, where mountains would impede mass attacks. The landing at Leyte saw one last throw of the dice by the Imperial Japanese Navy, which came within the barest margin of success. Admiral Halsey, in command of the US 3rd Fleet, was lured away in pursuit of the Japanese carriers, now neutered with the skilled pilots of the Kido Butai mostly dead. This left the San Bernadino straits open for a surface dash by a heavy fleet built around the superbattleship Yamato, an attack which was only parried due to the heroism of Admiral Sprague's Taffy 3 escort group. Bold torpedo attacks and a moment of irresolution by Admiral Takeo Kurita saved the cargo fleet and troops on shore when the Yamato and its escorts turned back, its mission unfulfilled.

The Phillipines still had to be taken, of course, but the islands were big enough for MacArthur to launch sweeping flanking attacks with the aid of further amphibious landings. Manila was a foretaste of what was to come. The Japanese fortified the Intramuros old town and unleashed an orgy of violence on the civilian population. Intramuros was comprehensively destroyed in urban warfare at immense human cost.

Iwo Jima and Okinawa, the battles to come, would be even worse. Those islands were small enough that the Japanese army could fortify every bit of high ground, creating a system of interlocking bunkers and tunnels essentially immune to bombardment. These islands were taken in a series of grinding frontal attacks with heavy casualties on the American side, and total annihilations on the Japanese side. For the Navy, constant kamikaze attacks introduced a new form of dread. It required minimal flying skills to dive at a ship, at least compared to the intricacies of a torpedo attack, and every defense had to be perfect. Japanese sources consistently over-estimated the success of kamikazes, and while they rarely sunk larger ships, a successful attack could kill dozens or hundreds of sailors. Carriers packed full of aviation fuel and munitions were especially vulnerable, and the Third/Fifth Fleet (the number switched when Halsey or Spruance took command) had to send units back for massive repairs.

The end of the war is inseparable from the B-29 strategic bombing campaign and the choice to use the atomic bomb. The initial stratospheric raids with the B-29 were ineffective, with the poorly understood jetstream scattering bombs. USAAF staff had concluded Japanese cities were extremely vulnerable to fire, and in a series of maximum effort blows, Curtis Le May's bombers burnt down six cities in as many raids, completely exhausting the stock of incendiary bombs in the theater. Aerial mining and a series of raids by the fleet carriers dealt further blows to the faltering Japanese war economy. The US submarine fleet was essentially withdrawn due to lack of targets. The first invasion, codenamed CORONET, was scheduled for November 1, 1945. All involved expected it to be Okinawa on a much more massive scale. In Toll's estimation, there was no explicit decision to use the atomic bomb. It had been built, and it might end the war. American principles, such as President Truman, were involved with the Potsdam conference and the fate of Europe. One of Truman's overriding concerns was ending the war without massive Soviet involvement, which would strengthen their postwar position in the Far East. And while he would grow into the office, in the first few months Truman lacked the self-confidence to fundamentally alter FDR's plans, even if he had disagreed with using the atomic bomb.

Toll delves deeply into the confusing issue around the Japanese surrender. The key barrier was that the formula of unconditional surrender might include the end of the monarchy, and the emperor had a divine status. Militarist leaders who had lead Japan to war had their own lives at stake, and many of them either committed suicide or were executed for war crimes by postwar tribunals, but the future status of the Emperor appeared to be the key issue. In the months prior to the invasion, Japan sent out tentative peace feelers through neutral countries, but they pinned their hopes on a Soviet-mediated peace, and Stalin and Molotov deliberated delayed negotiations in favor of a Soviet military resolution in Manchuria. Worse, the consensus style of Japanese decision making at the upper echelons of power was incapable of reacting quickly, or breaking deadlocks. The central council broke evenly on surrender versus national suicide in the face of American power, and in the end only the personal intervention of Hirohito resolved the deadlock in favor of peace, though by then it was too late for the victims of Hiroshima and Nagasaki.

As always, Toll is a master of naval history. His careful step through the chronology illuminates the pressure that decision-makers were under, and the brutality of the end of the war. This chronology is aided by deep dives into specific topics: press censorship, pilot training, the Japanese home front. In some details, this final volume is a step back in terms of the precision of the writing. Toll has some favored phrases (blue tail pipes, violet dawn, Admiral Turner's drinking) that are overused. And his description of land battle lacks the comprehensive quality of his writing on war at sea. But hey, the whole trilogy is over 1500 pages, and on a subject nearly as vast as the sea itself. As a whole, this is a new standard for excellence in military history.