<P> In the 2001 book Theodore Roosevelt, the U.S. Navy and the Spanish--American War, Wegner revisits the Rickover investigation and offers additional details . According to Wegner, Rickover inquired with naval historians, at the Energy Research and Development Agency, about Maine, after reading an article in the Washington Star - News in which its author, John M. Taylor, claimed the U.S. Navy "made little use of its technically trained officers during its investigation of the tragedy ." The historians, then working with the admiral on a study of the U.S. Navy's nuclear propulsion program, said they knew no details of Maine's sinking . When Rickover asked whether they could investigate the matter, the historians, now intrigued, agreed . Knowing of Rickover's "insistence on thoroughness," Wegner says, all relevant documents were obtained and studied . These included the ship's plans and weekly reports of the unwatering of Maine, in 1912, by the chief engineer for the project, William Furgueson . These reports included numerous photos, annotated by Furgueson with frame and strake numbers on corresponding parts of the wreckage . Two experts on naval demolitions and ship explosions were brought in . Since the photos showed "no plausible evidence of penetration from the outside," they believed the explosion originated inside the ship . </P> <P> Wegner suggests that a combination of naval ship design, and a change in the type of coal used to fuel naval ships, might have facilitated the explosion postulated by the Rickover study . Up to the time of Maine's building, he explains, common bulkheads separated coal bunkers from ammunition lockers and American naval ships burned primarily smokeless anthracite coal . With an increase in the number of steel ships, the U.S. Navy switched to bituminous coal, which burned at a hotter temperature than anthracite coal, and allowed ships to steam faster . However, Wegner explains, while anthracite coal is not subject to spontaneous combustion, bituminous coal is considerably more volatile . In fact, bituminous coal is known for releasing the largest amounts of firedamp, a dangerous and explosive mixture of gases (chiefly methane). Firedamp is explosive at concentrations between 4% and 16%, with most violence at around 10% . In addition, there was another potential contributing factor in the bituminous coal--this was iron sulfide, also known as pyrite, that was likely present . The presence of pyrites presents two additional risk factors . The first involves oxidation . Pyrite oxidation is sufficiently exothermic that underground coal mines in high - sulfur coal seams have occasionally had serious problems with spontaneous combustion in the mined - out areas of the mine . This process can result from the disruption caused by mining from the seams, or other processing, which then exposes the sulfides in the ore to air and water . The presence of pyrites in coal has been recognized to be self - heating . The second risk factor involves an additional capability of pyrites to provide fire ignition under certain conditions . Pyrites, which derive their name from the Greek root word pyr, which means fire, can cause sparks when struck by steel or other sufficiently hard surfaces . Before the use of flintlock guns, for example, pyrites were used to strike sparks to ignite gunpowder in an earlier model type gun, known as a wheellock . In the presence of combustible gasses issuing from the bituminous coal, the pyrites could therefore have provided the ignition capability needed to create an explosion . A number of bunker fires of this type had, in fact, been reported aboard U.S. warships before Maine's explosion, in several cases nearly sinking the ships . Wegner also cites a 1997 heat transfer study which concluded that a coal bunker fire, of the type suggested by Rickover, could have taken place and ignited the ship's ammunition . </P> <P> In 1998, National Geographic magazine commissioned an analysis by Advanced Marine Enterprises (AME). This investigation, done to commemorate the centennial of the sinking of USS Maine, was based on computer modeling, a technique unavailable for previous investigations . The results reached were inconclusive . National Geographic reported that "a fire in the coal bunker could have generated sufficient heat to touch off an explosion in the adjacent magazine (but) on the other hand, computer analysis also shows that even a small, handmade mine could have penetrated the ship's hull and set off explosions within ." The AME investigation, however, did note that "the size and location of the soil depression beneath the Maine' is more readily explained by a mine explosion than by magazine explosions alone"'. The team noted that this was not "definitive in proving that a mine was the cause of the sinking" although it did "strengthen the case". </P> <P> Some experts, including Admiral Rickover's team and several analysts at AME, do not agree with the conclusion . Wegner claims that technical opinion among the Geographic team was divided between its younger members, who focused on computer modeling results, and its older ones, who weighed their inspection of photos of the wreck with their own experience . He adds that the data AME used for its findings were flawed concerning Maine's design and ammunition storage . Wegner was also critical of the fact that participants in the Rickover study were not consulted until AME's analysis was essentially complete, far too late to confirm the veracity of data being used or engage in any other meaningful cooperation . </P>

Who was responsible for the uss maine explosion