By Commander Alan D. Zimm, U.S. Navy (Retired); David Dickson; Keith Allen; Capt. Christopher Carlson, USNR; Bill Jurens; Anthony Lovell; Stephen McLaughlin; Gorka L. Martinez Mezo; Nathan Okun; Lieutenant Colonel James C. O'Neil, U.S. Army (Retired); Jonathan Parshall; Richard Worth

April 2005

(In reference to A. Biache, P. Hsu, C. Lucas, and J. Rodgaard, pp. 18-22, December 2004 Naval History)

This article contends that three "proportionally spaced splashes" in the photograph were "rooster tails" generated when a midget submarine lost depth control and its propellers broke the surface. If this is the case, then plume "A," furthest to the left in the photograph, would have been initiated first; the middle plume "B" was initiated next; and the rightmost plume, "C," last. By extrapolating from measurements provided in the article, the distance from A to C is about 100 feet, and B to C about 50 feet. The height of C is about 50 feet, B about 25 feet, and A about 10 to 15 feet. These measurements are not precise, and are subject to the inaccuracies inherent in using the published photograph, dividers, and measurements mentioned in the article. Using this information, we can gain some ideas on the speed of the object that caused these events.

First, consider the authors' contention that a midget submarine caused these disturbances. In shallow water in a constrained harbor while lining up for an attack, a midget sub would operate at bare steerageway, or two to four knots. A submarine traveling at three knots would cover the 100 feet between A and C in about 20 seconds. Since plume A appears to have not yet totally collapsed, then the water in the plume must have gone up for about ten seconds, and fallen back for just short of ten seconds. Using Distance = .5 (Acceleration of Gravity) (Time squared) , that would make the required height of the plume on the order of 1,600 feet tall, or about 32 times as tall as the tallest plume shown on the photograph. This calculation makes questionable the idea these plumes were made by a broaching midget submarine at operating speed.

An estimate of the speed of the object can be had by looking at the heights of plumes B and C. If we assume C has reached near its peak height at 50 feet, and B also reached a peak of 50 feet and has fallen back down to 25 feet, the distance B has fallen (25 feet) provides an estimate of the time between A and B. Using the same formula, this time difference is 1.25 seconds. This estimates that the object creating the plumes was moving at about 40 feet per second, or about 25 knots. The maximum submerged speed of a Japanese midget submarine was 19 knots.

We tried calculations using different assumptions, and generally they indicated the (presumed) object was moving faster than 25 knots. For example, if the peak height of B was smaller than C (as is possible, considering the relative size of their bases), then B has fallen back less, making the time between the B and C less, and thus the speed of the object would be greater than 25 knots.

Doubling the speed of the submarine to six knots would require plume A to be in the air for more than ten seconds, requiring a 400-foot plume. A 50-foot plume would go up and collapse in about 3.5 seconds, requiring a submarine speed of more than 28 knots. Thus, the photographic evidence does not appear to support the idea these plumes were made by a submarine -- indeed, it appears to rule it out.

The article's authors developed a very convoluted argument to support their thesis. Occam's Razor would have us look to simpler explanations. From the above calculations, a more likely explanation was that the splashes were formed by a porpoising torpedo dropped by one of the attacking aircraft and in the process of accelerating to its rated speed of 41 knots.

Other evidence can be mustered to support this. First, there appear to be the marks of two recent torpedo drops on the water to the left of, and in line with, the plumes. Second, we know the Japanese had to modify their torpedoes to operate in the shallow waters of the harbor; we also know these modifications were not always successful, since some of their torpedoes malfunctioned during the live torpedo drops conducted during the air crews' training program. In the actual attack, the Japanese dropped 36 torpedoes (four torpedo bombers were shot down before they could launch), and we can account for between 19 and 22 that hit ships and exploded, so between 14 and 17 torpedoes malfunctioned or otherwise missed their targets.

Thus, a simple explanation for any presumed plumes of water on the photograph that is consistent with the data might be an aerial torpedo with defective depth control. This type of malfunction, called 'porpoising,' was common.

We say "presumed," because it cannot be conclusively proven the marks on the photograph are plumes of water. If they are plumes, they are curious, as they cast no shadows on the water that we could see. Alternately, the marks might be shell splashes, drifting smoke, AA shell bursts, or even just imperfections on the film.

In their Discovery Channel television special the authors provided a detailed look at their extensive effort to reproduce the "plumes" using a model of the midget submarine with her contrarotating propellers. The splash generated in their test looked nothing whatsoever like the "plumes" on the photograph; this they admitted in the program. Why the authors do not consider this conclusive evidence against their thesis was not explained. Indeed, it appears a double standard exists: the sub model broaching test was considered inconclusive, but a model test of an air-dropped torpedo was offered as conclusive evidence the plumes were not entry splashes from air-dropped torpedoes -- even when their tests and comparison photographs did not duplicate the conditions of dropping a Type 91 Mod 2 Japanese torpedo from a Kate torpedo bomber.

The authors claim this midget submarine launched its torpedoes against the U.S. battleships. This is key to their argument, because the launch supposedly caused the loss of trim in the submarine that, coupled with the shock waves from the detonation of the torpedoes, is what they claim caused the submarine's propellers to broach. We would like to point out, however, that the ten torpedoes carried by the five midget submarines all are conclusively documented: Midget A's wreck (sunk by the Ward DD-139 before the air attack), recently found, had both torpedoes on board; Midget B (sunk by the Monaghan DD-354 on the west side of Ford Island) launched both torpedoes, with the wakes observed by independent eyewitnesses; Midget C (washed up on Oahu outside Pearl Harbor) was captured with its torpedoes on board; Midget D wreck (found off Pearl Harbor in 1960) had both torpedoes on board; and Midget E launched both torpedoes against the cruiser St. Louis CL-49 near the harbor mouth.

The authors contend, without evidence, that the attack on the St. Louis did not occur. Instead, they argue that since all the other midget submarines either penetrated the harbor or were destroyed in the attempt, Midget E therefore must have penetrated the harbor, as they say, 'in the fabled tradition of the samurai.' This is in contradiction to the eyewitness account contained in the St. Louis's operations report, dated 10 December 1941, which states: "When just inside entrance buoy No. 1 two torpedoes were fired at this ship from a distance of approximately 2,000 yards on the starboard beam. The torpedoes, although running shallow, struck the shoal inside buoy No. 1 and exploded, no damage to this vessel resulting."

Could a submarine other than the Midget E have fired these torpedoes? The Japanese had stationed 22 submarines in the Hawaiian Islands area. Two, the I-16 and I-20, were stationed near entrance to the harbor where the attack occurred; three more, the I-24, I-22, and I-18, were stationed about 15 nautical miles from the entrance. According to their logs, none of these submarines made any attacks or expended any torpedoes that day.