These are excerpts from an article in Military Magazine Online by Colonel John H. Roush, Jr. Roush discloses that Japan had tested a nuclear device on 9 August on one of their remote islands and planned to use small nuclear warheads installed on kamikaze-directed bombs against invasion troop ships. [This is news to me and I haven’t indepentently verified it]Thanks to reader Scott, here’s a description of the Japanese atomic program from Answers.com:

After the war, the U.S. occupation forces found a total of five cyclotrons, which they judged to be part of the weapons programs. Cyclotrons can be used for electromagnetic uranium enrichment as mass spectrometers, but by themselves would not be useful as production facilities. In the United States, large cyclotrons at the Berkeley Radiation Laboratory were used to develop even more massive Calutron machines at the Oak Ridge facility, which were used for the bulk of the electromagnetic enrichment, but the Japanese cyclotrons would have been much smaller than even the prototype American machines. The Japanese cyclotrons were then dumped into Tokyo harbor by the U.S. Army, though many American scientists tried to intervene, insisting that the cyclotrons by themselves couldn’t be used to make atomic weapons.

[ . . . ]

Very little is known about the size of the atomic program in Konan though it is conventionally thought to have been small in comparison with the successful U.S. effort. In 1946, a journalist named David Snell working for the Atlanta Constitution wrote a sensationalist story which indicated that Japan had in fact successfully developed and tested a nuclear weapon in Konan. Snell was a former reporter, soon to become Life Magazine correspondent assigned to the 24th Criminal Investigation Detachment in Korea. He interviewed a Japanese officer who said he had been in charge of counter intelligence at the Konan project before the fall of Japan.

According to the officer, who used a pseudonym in the article because he was afraid of retalliation by occupation forces, the program was able to assemble a complete nuclear weapon in a cave in Konan and detonate it on August 12, 1945 on an unmanned ship nearby. Reportedly, the weapon produced a mushroom shaped cloud with a diameter of about 100 m (the first American bomb, “Trinity”, had a mushroom cloud some three times the size of that), and also destroyed several ships in the test area. To the observers 20 mi (32 km) away, the bomb was brighter than the rising sun. The officer then claimed that the Russian Army, which captured Konan in November 1945 after some of the last fighting in the war, dismantled the Japanese project and shipped it and some of its scientists taken prisoner back to the Soviet Union.

[ . . . ]

A 1985 book by Robert Wilcox reprinted the Snell interview as a basis for investigating the Japanese WWII nuclear efforts. In addition to detailing the known Japanese army and navy efforts, the book cites numerous intelligence reports and interviews which indicated the Japanese might have had an atomic program at Konan. It also gave evidence that the Japanese navy, taking up the atomic project after Nishina’s Riken had been destroyed, increased the Japanese efforts to make a weapon. The book, prefaced by Derek deSolla Price, Avalon professor of the history of science at Yale University, who endorsed it, was both panned and praised. Price wrote, “No longer can we maintain that a Japanese bomb just couldn’t have happened. Obviously it ‘nearly’ did. The only questions are how near and what does it do to our judgment on the one case we have of atomic warfare.” James L. Stokesbury, author of A Short History of World War II, wrote: “I had no idea the Japanese were working as seriously on an atomic bomb…this has to modify our perception of one of the crucial issues of the war.”

It is appropriate to ask why President Truman ordered the dropping of the bombs. In 1945, 85% of the people polled agreed with the decision, and only 10% disagreed that it was necessary to bring about the end of the war. If you had asked veterans of the war, I believe it would have been 100% in favor.

[ . . . ]

To understand [Truman’s] decision, consider OPERATION DOWNFALL, which once was labeled top secret . . . In the first invasion, codenamed OLYMPIC, U.S. forces were to land on Kyushu on 1 Nov 45. The second invasion, 1 Mar 46, codenamed CORONET was to involve 22 combat divisions against one million defenders on the main island of Honshu.{br />

[ . . . ]

By the summer of 1945 it was clear to analysts that Japan was losing the war, but intercepted cables disclosed that Japanese leaders were not considering surrender. There was no serious response to many approaches for a peaceful conclusion.

[ . . . ]

Senior military leaders of the day were in almost unanimous agreement that an invasion would be necessary to end the war. Those decisions were carefully considered based upon several imperatives:

• The hard-line rulers of Japan would not recognize defeat nor respond to peace overtures. Their troops would not surrender just by being overwhelmed in battle. The decrypts of intercepted messages disclosed this and the following points.
• We had learned lessons from the terrible battle of Okinawa that took place 350 miles south of Kyushu. Okinawa was defended by over 100,000 Japanese troops, and further supported by 50,000 home defense forces . . . The Okinawa invasion was comparable in size to the invasion of Normandy — 550,000 U.S. military personnel were involved and sustained more than 68,000 casualties; 34 allied ships were sunk, and 368 were damaged by 3,000 kamikaze planes. The Japanese demonstrated their willingness to fight to the last man. The Navy had more casualties in that campaign than the total sustained in all previous operations in all wars. [emphasis added]

[ . . . ]

• Japan was rapidly mobilizing for all-out defense of the home islands. Their never-defeated major army was brought back from Manchuria. The entire populace was being fully mobilized and systematically prepared to fight the invasion.

[ . . . ]

• Orders had been disseminated by the Japanese high command to kill all POWs, some 150,000 in Japan, plus 80,000 in other occupied Asian areas, plus interned civilians of enemy countries — a total of perhaps 450,000 — upon the initiation of an invasion attempt.

Following the capitulation, captured documents and interviews of key military persons disclosed the elaborate plans that had been made by the Japanese to defend their home islands. They had 2 million regular troops ready and 25 million additional men, women and children mobilized . . .

[ . . . ]

Japan tested a nuclear device on 9 August on one of their remote islands and planned to use small nuclear warheads installed on kamikaze-directed bombs against invasion troop ships. The Japanese military expected to kill 10,000 troops with each expenditure . . . [emphasis added]

The Japanese planners were convinced that they could shatter any invasion force so that the demoralized Americans would back off and they would accept a less than unconditional surrender on face-saving terms.

Even if the Allies gained a foothold on Kyushu, the Japanese had plans to mobilize the full population in the fiercest and most fanatical defense that could be conceived. Army officers had organized youngsters (the schools had been closed) into suicide brigades.

[ . . . ]

The Japanese had a national slogan: one hundred million will die for the Emperor and nation — and few doubted that they were, as a whole, prepared to fight to the death. Twenty-eight million had been enrolled in a National Volunteer Combat Force, some inadequately armed, but fully committed. We would have had to use our entire military strength in defense.

In the initial planning for the invasion it was thought that our casualties would be 150,000, a figure considered tolerable. However, that estimate had to be raised monthly. The final calculation exceeded one million. [emphasis added]

Clearly, the loss of lives would have been far greater than initial estimates, probably more than twentyfold, had the invasion plans been carried out. The nuclear bombings averted the need for the scheduled invasions, the results of which would have been far worse than contemplated.

[ . . . ]

It is clear from a study of the intelligence available and the plans our military prepared in light of then-available information of the intent and capabilities of the enemy that the battle that was to ensue would have become the most awful bloodbath in the history of modern warfare. Had the invasion been undertaken, it is estimated that the Japanese military deaths would have exceeded a million, with over 2 million additional casualties, if not more, plus many millions of civilian casualties also.
 [emphasis added]