Contact Us   |    Join   |    Donate


CDR John Alden, a rnbnarine veteran of World War II, is a prolific writer, most notable for his Tire Fleet Submarine in tire U.S. Navy. He is a frequent contributor to THE SUBMARINE REVIEW and is very respected for his thorough and thoughtful commentaries 011 WW II submarine actions.


The USS R-12 (SS 89) sank accidentally during exercises off Key West, Florida on 12 June 1943 with the loss of four officers, 24 regular crew members, 12 trainees, and two Brazilian Navy officers. The only survivors were the Commanding Officer, LCDR Edward E. Shelby; officer of the deck L TJG William D. Whetstone; helmsman Sydney H. Pool, S2c; lookout John Kapral, TM3c; and lookout Edward F. Zielinski, TM3c, all of whom were on the bridge al that time. A subsequent Court of lnquiry concluded that “the sinking was probably caused by flooding through an open torpedo tube due to inadvertent, thoughtless, or inattentive operation by a person or persons unknown now deceased,” or by failure of the interlocking device. (Opinion 21, p. 135 . This and subsequent references arc to page numbers in the court’s Record of Proceedings dated June 22, 1943 unless otherwise indicated).
To the best of my knowledge, the full story of the R-12’s loss has never been publicly disclosed. Based on a careful review of the Commanding Officer’s Action Report, the proceedings of the court, the testimony of witnesses, the plans and type of construction of the boat, and personal experience with submarines, I believe that the court failed to recognize the strong possibility of structural failure and improperly concluded that responsibility for the sinking was chargeable to one or more lost members of the crew.


The R-12 was one of many older submarines returned to service during the build-up prior to U.S. entry into World War II. Recommissioned on 16 October 1940, she operated out of New London, Coco Solo, St. Thomas, and Guantanamo Bay until February I 943. Following an overhaul at New London, she was ordered to Key West. Arriving on 11 May, she conducted exercises under the operational control of the Fleet Sound School along with other R-boats regularly stationed there, but was not formally transferred to Submarine Division 12 until 7 June, five days before being lost.


According to LCDR Shelby’s Action Report (SS-89 Serial 002, 16 June 1943), the boat had spent the morning conducting sonar training operations with USS CORAL (PY-I 5) and came to the surface at about noon in order to shift to an adjacent area and perform torpedo practice with the target vessel, EAGLE 56 (PE56). Lunch was being served and the morning watch section was in the process of being relieved. The surfacing procedure was proceeding normally when Shelby went to the bridge, followed shortly by the three enlisted men. At about 1202 the diving officer, LT Roger N. Starks, called up the hatch that Whetstone had heard a bumping sound on the starboard side of the forward battery compartment, and asked whether it had been heard on the bridge. Shelby had heard nothing, and after looking over the side reported that he could sec no sign of anything that might have scraped the hull.

Continuing the surfacing procedure, when he saw that the stern was well up Shelby ordered “secure pumping #2, ride the vents on #2 main ballast.” He then told LT Starks to pump #I ballast tank (the forward one) dry in preparation for riding the vents on that tank as well. At about 1212 Starks reported that suction on #I ballast tank had been lost. Noting that the bow looked a bit low, Shelby ordered that the tank be checked using the high pressure pump. At this point L TJG Whetstone came to the bridge to relieve the captain and Lt. Starks reported that the high pressure pump had lost suction on #I tank. Shelby check the trim forward and aft, saw that the bow was “about 2″ or 3″ low, which was normal as the ship was already riding the vents on #2 main ballast,” and ordered Starks to ride the vents on #I ballast tank, start the hull induction blowers, ventilate the batteries into the engine room, and go all ahead standard on the engines. The men on deck then shifted their attention toward fixing
the boat’s position and heading to the exercise area.

All still appeared normal when al about 1220 the collision alarm sounded and LT Starks shouted up the hatch: “Forward Battery Flooding.” Shelby immediately ordered both tanks blown but felt the boat start to tip down by the bow and sink so rapidly that the water had almost reached the coaming of the bridge hatch. Shouting down for Starks to close the lower hatch, Shelby slammed the upper hatch shut. He and Pool then stood on it to keep it from
blowing open, but were soon washed off the bridge. Shelby estimated that only about 15 seconds had elapsed from the sounding of the collision alarm until the bridge was completely under water. As he got clear of the bridge, he saw the boat go down at an angle of 75 to 80 degrees with the screws stopped. The survivors gathered together and at 1233 by the captain’s watch they observed an eruption of air bubbles, oil, and cork fragments that continued for about 45 minutes. They then helped one another stay afloat until picked up by the submarine chaser SC-449 after being in the water 5 hours and 35 minutes.


It was agreed by all concerned that the only possible causes of the sinking were flooding through a torpedo tube or by structural failure of the hull in the area of the forward battery compartment. The Judge Advocate’s questioning, however, focused most heavily on the torpedo tubes and went into exhaustive detail concerning the operation of the tube doors and the condition of the interlock mechanisms, which had been the subject of a Squadron 7 memorandum of May 7, 1943 titled Torpedo Tube Operating Gear that the R-12 had never received.( 18) The court also went into detail about an incident the previous day where a leaking torpedo tube drain valve had allowed water to enter the forward trim tank. It asked extensive questions about the state of mind of the torpedomen and the moral of the crew, based mainly on testimony by the boat’s yeoman that the leading torpedoman “wasn’t feeling very good” and “didn’t seem very happy” when the two had spoken the evening before the boat’s loss. (64) However, no other witness
supported that view. To the extent that the crew was in any way discontented, one member put it best: “I think that any man serving on an R-boat is discontented. He would rather be out in the war zone.”(73)

The torpedo tubes were obviously suspect, and the court’s focus on them was to be expected. However, it found no direct evidence of anything amiss with the tubes, the interlocks, or their operators. The strongest supporting evidence was that the outer door of number I tube was open when the boat surfaced, “as it had been used throughout the morning for firing water slugs.” (14, 32) Number 4 tube was also flooded and the outer door had been open two hours earlier, but the captain believed it was shut at the time of sinking. Tube 2 contained a ready war shot and number 3 was loaded with the exercise torpedo scheduled to be fired, after which it was to be reloaded with a second war shot; neither tube would have been flooded at the time. The officers and the surviving torpedomcn were unanimous in declaring it unthinkable that anyone on board would have opened both doors of a torpedo tube at the same time. The court’s conclusion that the sinking resulted

from flooding through the tubes rests only on the following :
Finding of Fact 32. “That the torpedo tube interlocking mechanism was in good condition, operationally and materially, so far as could be determined, but that this system is not absolutely fool-proof. lt can be easily disengaged manually. so that both tube doors can be opened at the same time,( 131) Opinion 11. “That one or two tubes’ outer doors may have been open at the time of surfacing.”( 133)


LT JG Whetstone’s report of hearing a thump in the battery room was very specific. He placed the noise between the air conditioning unit and the manhole cover to no. l main ballast tank on the starboard side and likened it as “similar to the letting out of the vacuum out of a tin can” or “the sound of a metal drum which had expanded in the sun.”(42) Two other men who also heard it and wondered what it was did not survive, but Whetstone was impressed enough to report the thump to the control room and then inspect the compartment for possible leaks. LCDR Shelby confirmed that it was relayed to him on the bridge, but the court made no further investigation of the possible source of the thump and did not mention it in its findings. Flooding was reported in the battery room, not the torpedo room. Testimony indicated that eight or ten men were probably in the torpedo room either on watch or eating lunch, and that some officers were probably eating in the forward battery compartment. LT Joseph G. Anthony, a naval architect, later testified that water from an open upper tube would have spouted about thirteen feet into the room and would have taken about 20 seconds to reach the door’s sill. Since the torpedo room was 31 feet long, presumably someone should have been able to report a flooding tube. However, the court concluded in Opinion 22: “That the water rushed through an open tube, through the torpedo room, cascaded into the forward battery room, and resulted in the report “Forward battery flooding.”( 135) Certain features of the surfacing procedure also pointed to possible structural problems. Both LCDR Shelby and LT JG Whetstone testified that the procedure used in the R-12 differed from that followed in other R-boats.(45, 123) They described the differences only generally as relating to the control of air to #I ballast tank, opening the drains, and getting ready on the engines.

The court made no effort to ascertain exactly what was done differently on other boats. It was mentioned that Shelby normally insisted that as little air as possible be used on surfacing, a reasonable measure to save precious compressed air. Riding the vents was not the peacetime practice on R-boats, but was undoubtedly done by the other boats at Key West. The court apparently felt that none of these procedures had any bearing on the sinking and did not go into them any further.

While no specific fault can be found with the general surfacing procedure, several possible problems were identified during the boat’s final surfacing. Whetstone, who was standing near the air inlet to #I main ballast tank, thought more air than normal was being used. Zeilinski, who was standing in the control room waiting to go up to the bridge, noticed that #I tank kept losing suction and venting air to the extent that LT Starks crossed the compartment to see what was wrong. Karpal was manning the stem planes and noticed nothing unusual before going to the bridge, where he heard Starks report that suction had been lost and the captain order using the high pressure pump because the bow looked a little low. After he was in the water he thought the boat went down with a starboard list.

The most significant testimony was given by Pool, who had been standing near the air manifold. He too noted that # 1 tank was spasmodically venting and taking suction and that much more air was being used than usual. Happening to glance at the high pressure air gauge, he thought the pressure dropped from 2200 to 1400 pounds per square inch. Pool had only been on the boat a month, and the Judge Advocate questioned him sharply: “Why would an inexperienced man notice such things as the dropping of pressure on a pressure gauge, the suction and venting of ballast tanks, which usually requires a qualified submarine man to observe and analyze?”(54) Pool responded that because he was fresh from submarine school he was trying to pick up knowledge about everything that was new to him, and reiterated his testimony when re-examined later. (94) The court dismissed the above considerations without comment, concluding in Opinion 15, “after witnessing surfacing under similar conditions in the R-4” that the air pressure drop was not considered abnormal.( 134) Actually, the test was made with the R-4 alongside the dock, in no way comparable to the conditions prevailing on the R-12.(114)

The nature of the final dive indicated that the R-l 2’s buoyancy was lower than normal. A II of the survivors testified that the boat seemed to drop out from under them before nosing down and was completely submerged in a matter of seconds. Shelby said that with the boat normally riding the vents it was practically impossible to get under in 45 seconds. LT Anthony testified that with the boat in trim and riding the vents its reserve buoyancy would have been reduced from 86 tons to 60 tons, and could have been as low as 40 tons if the tanks had not been completely dry before riding the vents. The capacity of the battery room was 52 tons, but even with 60 tons of reserve buoyancy, he said that “shifting of the water ballast in partially filled tanks and compartments would, in all probability, put the bow under.”(98) Asked specifically: “What volume of water entering the forward battery room would cause the bullnose to go under,” he responded: “If flooding resulted from a ruptured tank top, no water would be required in the battery compartment as number one main ballast tank would be flooded.”

There were clearly weaknesses in the structure of #I main ballast tank, which was of riveted construction. The tank had been pressure tested at New London, but parts were inaccessible for inspection. These included the forward bulkhead which was obscured by the air bottle well, the area under the magazines, the entire part constituting the battery well, and part of the port tank top that was under lockers. LT Anthony was asked to explain how a riveted joint might fail under pressure. He said a lap riveted scam would probably fail by gradual spreading to “an ever-increasing rupture” and that old rivets in a weakened plate might slip out with little or no audible sound.( I 00)

Shelby was more concerned that when the air conditioning unit was installed, holes had been cut in the starboard tank top and outer hull for the cooling water pipe and doubler plates had been welded around the openings. He suspcclcd that the welding might have set up some stress in the plating and questioned Anthony about this. The naval architect replied: “If the inside of the tank were subjected to negative pressure, atmospheric pressure on the other side of the plate would cause deflection against the negative pressure. Pressure in the tank would cause deflection in the opposite direction. With a poorly welded doubler plate it is quite possible that the weld would crack due to these alternate variations in stress.” ( 101) Shelby then asked whether this could cause the plate itself to rupture, and Anthony answered that it would not, because of the increased thickness of the doubler plate. No further questions were asked and the court recessed, with the result that the possibility of a riveted joint’s failure under repeated fluctuations of pressure was never explored. The court ultimately dismissed the possibility of structural failure in Opinion 20: “That, by the process of elimination, the loss of the R-12 cannot be attributed to: ” … (i) Shell plating” or, among other things, “Failure of number one main ballast tank top or bulkhead.” ( 135)


In addition to its effort to discredit Pool’s observation of the air pressure gauge, there are other indications that the court had prcconcluded that fault lay with the torpedo tubes. Although the survivors were unanimous in favoring structural failure over flooding through a torpedo tube, the Judge Advocate seems to have gone out of his way to shake LCDR Shelby’s conviction in the following exchange:

Kelly: “Having been at 90 feet, it is almost conclusive that the boat was structurally sound and that if any weakness had been present, the failure would have taken place while at this depth. Do you agree that this eliminates a structural failure?”
Shelby: “No.”
Kelly: “Why not?”
Shelby: “I have no evidence to convince me that it couldn ‘t have been a structural failure as well as it could have been a torpedo tube.”
Kelly: “What type of structural failure do you believe could cause this large volume of water to enter forward’?”
Shelby: “It would have to be a direct failure of either the hull or the sides of the ballast tank, and a large failure to admit the quantity of water necessary to cause the boat to go down as she did. However, I have no evidence that this occurred.”
Kelly: “Any structural failure of the nature you mentioned would no doubt have taken place at 90 feet rather than on the surface. Do you agree to that’?”
Shelby: “No sir, I do not agree. I would not make a definite statement either way, that it would or it wouldn’t. “(30-31) Following the testimony of the naval architect, CDR W. W. Weeden, Jr., Commander Submarine Division Twelve, was called to testify. A )though he stated that the R-12 had only been in the area for a month and under his command for five days, he considered that the boat’s material condition was good and that “the R-12 was organized and trained under the same standardizations as other R-boats(.)” (I 06) The court then asked his opinion as to what caused the loss of the R-12, to which he replied “that in view of the fact that no noise was heard indicating a hull rupture, the water must have come through a torpedo tube. This would presuppose that the outer and inner door of the tube were open at the same time.”(109)

On the sixth and last day of the hearings, LT William L. Fey, Jr., skipper of the R-4, was sworn in and asked by the Judge Advocate what he thought, as officer of the deck surfacing a boat under the conditions existing on the R-12, was the probable cause of her sinking. He replied: “My first impression would be a torpedo tube derangement. .. No other cause is conceivable to me.” (119) The stated conditions given him did not included mention of the thump reported by LT JG Whetstone. Asked by the court his opinion as to the possibility of structural failure, he responded: “I can think of no structural failure that would occur on the surface which would not already have occurred at 90 fcet.”(J I 9) The next witness was CMoMM Boyce Paul Mays, also from the R-4. He was asked, considering himself as chief of the watch in the control room of an R-boat under the conditions existing in the R12, what he thought would be the cause of the boat’s sinking, “no jar or shock having been felt.” He replied: “I would say something went wrong in the forward battery.”(119) Asked to explain what possibly could go wrong in that compartment, he cited a failure of part of the main ballast tank. Both the Judge Advocate and the court pressed him with further questions in an apparent effort to get him to concede that flooding could only have come from the torpedo tubes, but he would not agree.

None of the above three witnesses had heard any of the testimony during the inquiry, although CDR Weeden stated that he had read Shelby’s report of the sinking. Despite the obvious omissions of the thump heard in the forward battery room, neither LCDR Shelby nor L TJG Whetstone raised any objection or asked further questions of any of these witnesses.


While this issue was peripheral to the cause of the R-12’s Joss, it was within the charge to the Court of Inquiry. In peacetime, the fact that the survivors were left in the water for five and a half hours would likely have received extensive criticism in the popular press, but the court disposed of it rather perfunctorily. In its final
Finding of Fact it stated that a search was instituted “soon after the sinking.” (132) Actually, no one had become concerned about the R-12’s safety until about three hours after the boat had sunk. CDR Weeden, the division commander, testified that he was told at 1415 that the torpedo runs had been delayed and that EAGLE 56 was unable to contact the R-12. This did not particularly concern him, because radio communications in the area were often unreliable, but when the R-2 returned at about 1510 and reported that they had not seen the R-12 all afternoon, he tried unsuccessfully to reach EAGLE 56. At 1540 he telephoned the local fleet air wing and asked that a search be started. He also checked with the skipper of CORAL and learned that the R-12 had last been seen diving. At this point he reported to Captain Edward H. Jones, Commander Task Group 23.3 (the ships attached to the Sound School) that the boat was missing, the SC-449 and PC-451 were ordered to make a surface search. At I 722 he notified ComSubLant that the R-12 was missing and probably sunk. Other ships and planes were sent out, and at about 1750 a circling plane pointed the SC-449 to a large oil slick and a lookout spotted the survivors in the water.( I 02- I 03) LCDR J. H. Church, Jr., in Command of USS CORAL, testified that his orders were to escort a submarine back lo the base except on Saturdays when the boat was staying out for torpedo exercises.

He was not instructed to tum the submarine over to another escort, but once it had surfaced he was free to return to port. The ships were about a mile apart when he saw the R-12 dive, noticed nothing abnormal about it, and did not sec men in the water. Assuming the submarine had made a practice dive, he continued back to the base.
Captain Jones confirmed that an escort was usually allowed to return to port early on Saturday when “(t)here is always a submarine officer on board of the surface ship conducting the purely submarine exercises which have nothing to do with Sound school.”( 116)

The ship in question was the EAGLE 56 and the officer conducting the practice torpedo firings was none other than LT Kelly, the Judge Advocate of the Court of Inquiry. Calling himself as a witness, he asked himself to state all he knew about the casualty. He said that there had been about an hour’s delay in getting the radio operating, because of which only the R-2 was contacted immediately to make the first torpedo run. Once this had
begun, he tried to contact the other boats, and when the R-12 failed to respond he moved the R-13 and R-11 up to its place. Between about 1255 and 1630 continual efforts were made to contact the R12 without success, but Kelly felt no concern because he knew the boat was operating with a surface vessel escort that would have known of any problem. He asked the R-11 if she was able to reach the R-12 by radio, but received a negative reply and did not notify anyone else of his inability to contact the R-12, assuming that the submarine had broken down or returned to base for some other reason. (5-6) The court asked LT Kelly no further questions of its own.

In Opinion I 8 the court declared that the system for accountability of submariners at Key West was satisfactory except on Saturdays when they were “changing from sound exercises to torpedo exercises. The accountability during this period should be more positive.” (134) That final sentence should have been stated as a positive recommendation, but the court saw fit to play it down. Indeed, a more stringent finding would have reflected adversely on many of the officers involved, and might even have led to charges of negligence. Captain Charles F. Erck, the president of the Court of Inquiry, was also ComSubRon Seven and thereby the direct superior of CDR Weeden, ComSubDiv Twelve. LT Kelly, the Judge Advocate, was on the SubDiv staff and had been in direct
control of the torpedo firing exercises. All three were therefore interested parties rather than neutral investigators.


Several defects in the court’s conclusions were recognized in the review process. Rear Admiral Freeland A. Daubin,
ComSubLantFlt, the convening authority, noted that Opinion 21 (b) was misleading in implying that flooding could have been caused by mechanical failure of the interlock, whereas “the doors would have to be opened physically by a person.” He disagreed with Opinion 25 that placed responsibility for the boat’s loss on a person or persons unknown, “as there is no basis for such an assumption.”

The sinking could only be attributed to sudden massive flooding, but “the source of the entry of the water cannot be determined except by conjecture.” (Sublant SECRET attachment, 9 July 1943) The reviewing authority, Admiral R. E. Ingersoll, CinCLantFlt, stated that “it is not believed that the possibility of structural failure can be eliminated” and “that the sinking could be attributed to any of the causes’ that had been ruled out in Opinions 20(1), (j) (m), and (s) as specific possibilities of structural failure. (LantFlt SECRET hr serial 00848 of 17 Jul 1943) Admiral Ernst J. King, ComlnChUSFlt and CNO, approved the proceedings “as modified by the comments of the convening and reviewing authorities,” thereby effectively clearing the crew of responsibility for the loss of their ship but left the cause undetermined as between flooding through a torpedo tube or a rupture in the hull structure. (ComlnCh SECRET hr serial 001463 of 22 July 1943) Inspection or salvage of the wreckage was impossible al that
time, other wartime problems had higher priority, and no further investigation was ever made.


In my opinion the case for structural failure is much stronger than was found by the Court of Inquiry. Key members of the court, possibly because of certain conflicts of interest, focused on a cause that implied error on the part of R-12 personnel and tried to divert attention from other conditions that might have reflected unfavorably on themselves. The Judge Advocate in particular seems to have acted more like a prosecuting attorney than a neutral investigator.

The R-12 was something of an orphan boat; it had not been at Key West long enough for superior officers to have become fully acquainted with its personnel or its material condition, or for its own officers to adapt completely to local procedures and customs. In concluding that flooding had occurred through a torpedo tube, the court gave greater credence to the opinions of R-4 personnel based on hypothetical assumptions than to the repeated assertions
of the R-12 survivors.

None of the members of the court had expertise about possible modes of structural failure. A It hough a well-qualified naval architect was available at Key West, he was not asked to evaluate the complete evidence and the defendants were not knowledgeable enough to ask him the essential searching questions. In my opinion this constituted a major deficiency in the inquiry. It may be purely coincidental that two other boats of similar design and riveted construction – the 0-9 on 19 June 1941 and the S-28 on 4 July 1943 sank with all hands during submerged operations, from causes that were never determined. The survivors of the R-12 provided clues that may have a bearing on all three sinkings.

I am not a naval architect, but l have an engineering background. At submarine school 1 made training dives on the R-9 and the similar 0-4 and 0-8. 1 served on five fleet boats and am qualified in submarines. Later I spent several years as an engineering duty officer working in shipyards on submarines and other ships. On the basis of that experience, 1 believe the following plausible hypothesis can be made for the loss of the R-12.

The forward ballast tank was complicated in form and had many riveted joints, several of which were in locations inaccessible for regular inspection. Since the boat’s recommissioning the tank was subjected to alternate pressure and suction during 519 dives. The thump heard by L TJG Whetstone was characteristics of a joint springing apart, which would have allowed water to enter and air to vent out of the tank. This probably started slowly but was
noticeable enough that use of the high pressure pump was called for, which would have produced a greater vacuum than normal and tend to widen the gap in the joint. The wartime practice of riding the vents reduced the boat’s reserve buoyancy by 30 to 50 percent and also accustomed the bridge personnel to seeing the bow ride lower in the water. When the kingstons were opened to ride the vents, the forward main ballast tank continued to fill gradually until water finally entered the forward battery compartment through either the battery well, a manhole, or some other opening. By the time it was first noticed and reported, the buoyancy had been reduced to the point that the boat started lo go under immediately. The free surface effect of water in the partially filled spaces then
caused the bow to tilt down rapidly and the boat to sink at a steep angle.


LT Starks was posthumously awarded the Navy and Marine Corps Medal, while Stanley Pool received a Letter of Commendation by the Secretary of the Navy and was recommended for a Life Saving Medal for keeping LT JG Whetstone afloat while awaiting rescue. LCDR Shelby went on to take command of SUNFISH (SS28 I) and conducted five war patrols in the Pacific. Whetstone was ordered to the new construction submarine PIPEFISH (SS388) and served in the Pacific. The loss of the R-12 must have preyed on his mind to the extent that he confided his concerns with shipmate Frank Ferguson, another R-boat veteran and a lifetime friend until Whetstone’s death in 1970. The two apparently concluded that the suction of the high pressure pump created a heavy vacuum in the tank, producing a break that allowed the tank to refill and flood into the forward battery. Y cars later Ferguson sent this information to the National Archives and wrote a short article about it in Polaris magazine (Polaris/December 1999, p. 15). It was this article that stimulated my interest in the case.

Naval Submarine League

© 2022 Naval Submarine League