Footnotes

[1]. See, e.g., Edwin P. Hoyt, The Lonely Ships: The Life and Death of the US Asiatic Fleet, New York, David McKay, p. 257; W.G. Winslow, The Ghost That Died At Sunda Strait (“Ghost”), Annapolis, Naval Institute Press, 1984, p. 125.

[2]. J.A. Collins, Commodore, RAN, “Reports on the Battle of the Java Sea,” in Ronald McKie, Proud Echo, London, Robert Hale, 1953, p. 135.

[3]. See, e.g. Winslow, The Fleet the Gods Forgot (“Fleet”), Annapolis, Naval Institute Press, 1982, p. 209 and Ghost, p. 123; “Partial Log As Kept By Survivors, USS Houston,” enclosure (a)(9), 9 September 1945 (found at Hyperwar: World War II on the Worldwide Web (“Hyperwar”): http://www.ibiblio.org/hyperwar/.

Note that nothing in this article should be interpreted as criticism of the survivors of the Houston, who had to endure three years of brutal treatment in Japanese POW camps before they could be interviewed by Allied authorities.

[4]. H.M.L. Waller, Capt., RAN, “Action Narrative B Day and Night Action off Surabaya, February 27, 1942,” in McKie, p. 139.

[5]. Also spelled “Surabaja,” “Soeurabaya,” “Soerabaya” or somesuch.

[6]. Look, for the last time it’s pronounced “CHIL-a-chap,” OK?

[7]. Such casualties included, in no particular order:

1. light cruiser USS Boise struck an uncharted reef in the Sape Strait and was lost to a planned ABDA counterlanding operation off Balikpapan;
2. light cruiser USS Marblehead blew out an engine and was lost to the same ABDA counterattack off Balikpapan;
3. destroyer USS Whipple collided with De Ruyter in a fog;
4. destroyer USS Edsall dropped a depth charge at too slow a speed and it detonated under her stern;
5. destroyer Hr. Ms. Van Ghent ran aground in the Stolze Strait and had to be scuttled;
6. destroyer Hr. Ms. Kortenaer lost rudder control and ran aground off Tjilatjap and was thus lost to the ABDA counterattack off Bali;
7. destroyer USS Stewart rolled over in drydock in Soerabaja and was scuttled (ineffectually, as it turned out, as the Japanese were able to salvage her);
8. destroyer USS Pope developed a leak in the feed pipes to her boilers and was unavailable for the action in the Java Sea;
9. destroyer Hr. Ms. Witte de With had one of her own depth charges detonate under her stern; and
10. destroyer HMS Jupiter was sunk after she apparently struck a discarded Dutch mine.

[8]. See, e.g., Duane Schultz, The Last Battle Station: the Saga of the USS Houston, New York, St. Martin’s Press, 1985, pp. 109, 117-118.

[9]. Doorman has been subject to severe criticism for his performance in the Netherlands East Indies campaign. Morison obliquely criticized Doorman’s caution. Samuel Eliot Morison, History of United States Naval Operations in World War II, Volume 3: The Rising Sun in the Pacific, Edison, NJ; Castle, 1948, pp. 310, 311, 340. Morison also notes specifically that the US officers lacked confidence in Doorman. Id., p. 338. John Prados called Doorman “aggressive to the point of recklessness.” John Prados, Combined Fleet Decoded, Annapolis, Naval Institute Press, 1995, p. 255. Edwin Hoyt notes that Doorman was criticized as “misguided and stubborn.” Hoyt, p. 258. Mike Coppock has probably been the most brutal:

As he yelled "Follow me!" in one of the most desperate and ill[-]conceived sea battles in modem times, an ad hoc fleet of Allied warships was unnecessarily squandered in a do-or-die encounter that became the Armageddon of ego-driven Dutch Admiral Karel Doorman. (Mike Coppock, “The Battle of the Java Sea: A Fleet Wasted,” Sea Classics, Sept. 2007.)

While a full analysis of Karel Doorman’s performance is beyond the scope of this article, I must put myself on record as finding most of these statements more than a little unfair. In my opinion, while he did make mistakes (for instance, he tried to get “too cute” in action off Bali), Doorman’s actions in the campaign as a whole and in the Java Sea battle in particular were at worst defensible and generally tactically sound. That he was unsuccessful is not so much a reflection on his ability as it is on the situation that was forced upon him, in which he was left with numerous judgment calls with few clear correct answers. I hope that some of the information I provide in this article will help provide a more balanced view of this brave, humane and honorable officer.

[10]. So named, actually, because it was the combination of the Western Striking Force from Tanjoeng Priok in western Java and the Eastern Striking Force from Soerabaja in eastern Java.

[11]. Waller’s report in McKie, p. 142.

[12]. ONI Narrative (found at Hyperwar), p. 53.

[13]. Helfrich signaled, “Notwithstanding the air attack you are to proceed eastward to search for and attack the enemy.” ONI Narrative, p. 55. Not often reported is a second part of this message, “Air attacks had been expected and this attack should not have been a reason for withdrawing from the area of action.” Morison, p. 340. One wonders what went through the mind of Admiral Doorman, in danger at sea, upon receiving this message from a superior sitting safe at a desk in the mountains of Java. In that context, Doorman’s reply was a model of restraint, “Was proceeding eastwards after search from Sapoedi to Rembang. Success of action depends absolutely on getting good reconnaissance information in time, which last night failed me. Destroyers will have to refuel tomorrow.” ONI Narrative, p. 55.

[14]. ONI Narrative, p. 55.

[15]. This rather colorful anecdote is not generally mentioned. J. Daniel Mullin, Another Six-Hundred, Mt. Pleasant, SC; J. Daniel Mullin, 1984, p. 214.

[16]. The 4th Destroyer Flotilla consisted of the flagship cruiser Naka, serving as a destroyer leader as most Japanese light cruisers did, destroyers Asagumo and Minegumo, of the 9th Destroyer Division, and destroyers Murasame, Samidare, Harusame and Yudachi of the 2nd Destroyer Division. F.C. van Oosten, The Battle of the Java Sea, Annapolis, Naval Institute Press, 1976, p. 42.

[17]. The 2nd Destroyer Flotilla consisted of the flagship light cruiser Jintsu, destroyers Yukikaze, Amatsukaze, Tokitsukaze and Hatsukaze of the 16th Destroyer Division, and destroyers Yamakaze and Kawakaze of the 24th Destroyer Division. Destroyers Ushio and Sazanami of the 7th Destroyer Division, which had been operating as a screen for the 5th Cruiser Division, were attached to the 2nd Destroyer Flotilla for this battle. Van Oosten, p. 43.

[18]. Eric Lacroix and Linton Wells II, Japanese Cruisers of the Pacific War (also affectionately known as “The GIANT Book of EVERYTHING You Could Possibly Want to Know About World War II Japanese Cruisers But Were Afraid To Ask”), Annapolis, Naval Institute Press, 1997, pp. 245-250. CombinedFleet.com compares the Japanese Type 93 torpedo’s specifications to those of the US Mark 15 (used by US destroyers) as follows:



It should be noted that early in the war US torpedoes were almost completely ineffective due to a combination of an erratically-performing magnetic detonator, a defective firing pin and a miscalibrated depth-setting mechanism. These problems were not corrected in full until 1944.

[19]. Doorman has been subject to severe criticism for this decision, both by historians, naval analysts and even the crews of the Anglo warships. See, e.g., Mullin, p. 229. In fairness to him, however, it must be restated that he was acting in accordance with Allied doctrine in leaving his float planes ashore for an expected night battle. Unless a float plane could be stowed in a hangar – neither De Ruyter nor Java had hangars – it was regarded as a fire hazard. Tom Womack, The Dutch Naval Air Force Against Japan: The Defense of the Netherlands East Indies, 1941-1942, Jefferson, NC and London, Macfarland, 2006, p. 197 n.23.

The dangers of such float planes was exemplified the following August in the Battle of Savo Island, when a Japanese cruiser force made a surprise night attack on Allied cruisers off Guadalcanal. The Japanese set fire to exposed US floatplanes, which served as points of aim in the darkness. Four Allied cruisers were lost in this action.

[20]. Allied floatplanes were less capable than their Japanese counterparts by an order of magnitude and were never fully integrated into Allied tactics like their Japanese counterparts were.

[21]. Womack, p. 124.

[22]. Doorman was aware of the air issues and had arranged for US air officers to come aboard to act as liaisons for a flight of US B-24 Liberators to serve the Combined Striking Force. Why he wanted this arrangement over the use of his own floatplanes is unclear, but it may have been because of the Liberator’s superior range, armor and armament. The US air commander agreed to the plan but was overruled by ABDA-Air. Van Oosten, p. 39.

[23]. On the other hand, given the almost total ineffectiveness of US torpedoes early in the war, the decision may have actually made little difference.

[24]. Java’s sister ship Sumatra was indeed deemed unfit for service due to problems with her engines, which considering the dire straits in which the Dutch found themselves is saying something.

[25]. Mullin, p. 224.

[26]. Lieutenant (junior grade) Harold S. Hamlin, Jr., of Houston likened the situation to eleven all-stars playing the Notre Dame football team without a single practice session together. Toland, But Not In Shame, p. 244. At this point in time, Notre Dame was a national football power.

[27]. Prados, p. 261.

[28]. Hara Tameichi, Japanese Destroyer Captain, Annapolis, Naval Institute Press, 1967, pp. 65, 68-70.

[29]. “Crossing the T” describes a maneuver dating to the Age of Sail in which one or more ships cross in front of and perpendicular to one or more opposing ships, which allows the crossing ship to fire all of its guns (or all of its side guns in the Age of Sail) at the target while the target could only fire back with its forward guns (or none of its guns in the Age of Sail). In the case of the Battle of the Java Sea, Nachi and Haguro would have been able to fire their combined twelve forward and eight aft guns, for a total of twenty guns, against, theoretically, the three forward guns of De Ruyter, four forward guns of Exeter, six forward guns of Houston, four forward guns of Perth and four guns of Java, for a total of 21 guns. Doorman’s response here has been heavily criticized by historians. However, it should be pointed out that because the cross of the T formed by Nachi and Haguro would have been so short, the ranges of all ABDA guns except the three forward guns of De Ruyter would have been fouled by other Allied ships and, consequently, would be unable to fire.

Prados also points out that had Doorman kept closing the range he would have risked letting Takagi get behind him, in which case the Allies would have been unable to catch up because they were slowed down by the damaged Kortenaer. Prados, p. 260.

[30]. Kroese was popular with the Americans, to whom he was known as “Cruiser.” John Toland, But Not In Shame: The Six Months After Pearl Harbor, New York, Random House, 1961, p. 254.

[31]. ONI Narrative, p. 68. Some versions state the message was, “British destroyers counterattack.” ONI Narrative, n. 56.

[32]. ONI Narrative, p 78, Ph.M. Bosscher, De Koninklijke Marine in de Tweede Wereldoorlog, Franeker, Netherlands; Uitgeverij Van Wijnen, 1990, p. 286 n. 321.

[33]. Mullin, p. 226.

[34]. There has been some question concerning Binford’s authority to withdraw his destroyers, but a closer inspection shows that the action was correct. Doorman had indicated he only needed the US destroyers, which were otherwise underarmored, undergunned and slow, for their torpedoes. With the torpedoes expended – and especially with fuel almost gone – the destroyers were now effectively useless to the Combined Striking Force. Doorman’s comments during the DesRon 58 attack showed he was aware of their fuel issues. Binford sent a message to Doorman through Naval Commander Soerabaja informing the admiral of his withdrawal. Doorman responded by affirming the decision and ordering DesRon 58 to Tanjoeng Priok for new orders and a new load of torpedoes.

[35]. Mullin, p. 226.

[36]. Winslow, Ghost, p. 122, Fleet, p. 208.

[37]. ONI at 74. Morison, p. 356, has the signal as “I am torpedoed.”

[38]. Mullin, p. 226.

[39]. One survivor of De Ruyter said that the cruiser actually “stopped” to pick up the survivors of Jupiter. Mullin, p. 226. I have not been able to corroborate this statement.

[40]. That night the Allied crews were not convinced the culprit was a submarine torpedo; postwar research would reveal no Japanese submarine in the area that could have been responsible for the explosion. Allied suspicions instead focused on the Dutch minefield in Toeban Bay. It may very well be that Jupiter used the term “torpedoed” merely as shorthand.

The explanation that has been accepted is that one of the mines had broken loose and struck Jupiter. A “drifting Dutch mine” appears to be the most popular description. The most recent scholarship suggests, however, that while that description is not far from the truth, it does not seem to do this bizarre incident justice.

The minefield, it seems, was never laid. En route to the location, Gouden Leeuw was spotted by one of the ubiquitous Japanese float planes. Convinced an air attack was imminent – Japanese float planes were often used in air attacks B – Gouden Leeuw apparently dumped her volatile cargo, only some of which were armed, and left the area. Van Oosten, pp. 68-69; Womack, p. 197 n. 33.

[41]. Schultz, p. 158. His is the only source to give an exact course. Most sources only say the column headed “north” or “in a northerly direction.” The ONI Narrative seems to assume a course of due north but does not explicitly say so. Other sources dispute the course. Mullin, p. 226, says the column turned northwest. Morison has the column headed north northwest for the remainder of the action, but does not give a course. Morison goes on to say that when Takagi’s cruisers engaged Doorman later that night, they were “almost parallel,” but all sources seem to agree that the Japanese were headed due north. Van Oosten has them turning due north then at some unspecified point in time turning slightly northwest, but again does not give a course. Paul S. Dull, A Battle History of the Imperial Japanese Navy (1941-1945), Annapolis, Naval Institute Pres, 1978, p. 85, uses van Oosten’s figures. I am using Schultz’s figures, backed up by the ONI Narrative, mostly for simplicity, but keep in mind that the column was also zigzagging. Earlier in the day they had been zigzagging 10 degrees from the base course. “Partial Log As Kept By Survivors, USS Houston,” Enclosure (a)(11) “Engagement off Soerabaja, February 27, 1942,” (found at Hyperwar). It is possible that the zigzagging accounts for the reported differences in course.

[42]. On the morning of March 1, 1942, Encounter, along with Exeter and the destroyer USS Pope, would be chased down and sunk by the Japanese cruisers Ashigara and Myoko.

[43]. Kroese, pp. 89-90. Some translations use “light” instead of “flare.”

[44]. This story would seem to conclusively show the order of the Allied cruisers as they passed by the survivors: “The Dutchman, the Australian, The American and at last another Dutchman.” By this description, the front-to-back order would be De Ruyter, Perth, Houston and Java. The survivors were in the perfect position to watch the ships as they passed by, one by one, in the hope that one of them would stop. So it would seem that they conclusively determine that the column was in that order, unless they were mistaken in the identities of the ships.

Kroese’s excellent work hints at the possibility of such a mistake. The Dutch Navy at War is illustrated with a few drawings by H.J. Hoowij. One of these drawings shows the Allied cruisers passing by the Kortenaer survivors. The caption describes the order of Allied ships as De Ruyter, Perth, Houston and Java. Yet the picture itself appears to show the second cruiser in the column having a tall, tripod mast, which was a characteristic not of Perth but of Houston. Were the Kortenaer survivors mistaken as to the identity of the cruisers?

The strong likelihood is that they were not. The Dutch sailors were familiar with Houston both as one of the most powerful ships in their little fleet, a ship that had been with them in battle, and, along with De Ruyter and Java, the most visually distinctive. A hint is not enough to overcome the plain words of the witnesses.

[45]. ONI Narrative, p. 75.

[46]. James D. Hornfischer, Ship of Ghosts, The Story of the USS Houston, FDR’s Legendary Lost Cruiser, and the Epic Saga of Her Survivors, New York, Bantam, 2006, p. 91.

[47]. ONI Narrative, pp. 69-70.

[48]. Id.

[49]. Van Oosten, pp. 72-73, Womack, pp. 126-127. There has been considerable confusion on this point because while De Ruyter’s TBS radio went out, other parts of the Dutch cruiser’s radio suite apparently remained operational. At two points during the evening, Doorman was able to send short-wave, non-voice messages to Dutch shore installations. One was to Helfrich: “Enemy retreating westward. Contact broken. Where is convoy?” The second was in response to DesRon 58's report that they had retired to Soerabaja. Unable to reach Doorman on the TBS, they relayed a message through Naval Commander Soerabaja, which in turn relayed Doorman’s acknowledgment of their report and new orders for them. But for battlefield communications, the non-voice radio was impractical and even dangerous. Essentially, for the remainder of this action, the Combined Striking Force’s flagship had no radio.

[50]. Mullin, p. 226.

[51]. Van Oosten, pp. 72-73, Womack, pp. 126-127.

[52]. Collins’ report, p. 138.

[53]. Waller’s report makes no mention of such a signal from Doorman, but such a signal had to come. Waller simply did not have the authority to order another ship to pick up the survivors in contravention of Doorman’s order that such survivors were to be left “to the mercy of the enemy.” Additionally, Waller stated that he did not recognize the survivors or their language. Finally, Waller’s report makes no mention of Perth sending out a message to pick the survivors up, but it is known that such a message indeed went out. The radio message was sent from Perth, but the order originated with Doorman, not with Waller.

[54]. The radio log of John D. Ford paraphrases a voice message as “...pick up survivors we just passed in a boat.” Report of John D. Ford (228): “Report of Battle of Java Sea, forwarding of: Extracts from radio log 26 February to 1 March 1942.” Enclosure (B). The sender is unidentified; the intended recipients were listed as two British ships, which Bosscher identifies as, oddly, Electra and Jupiter. Bosscher, p. 612 n. 379. There was undoubtedly more to the message, giving at least the Kortenaer survivors’ location. With the voice radios on De Ruyter and Houston out of commission and Java at the other end of the column, this voice message could only have come from Perth.

[55]. Interestingly, Doorman’s original signal to Perth may have only been to have the survivors picked up without specifying who should pick them up. The message from Perth was directed to Electra and Jupiter. Doorman knew both were out of the battle; Waller knew about Electra but not Jupiter. So whoever sent the message did not know that Electra and Jupiter were out of the battle, which also indicates that the message was written by someone in the radio room who did not have access to the battle situation. It is not known if Encounter responded to Perth’s signal with a message of her own; Ford’s radio log does not indicate one, but the log by its own admission only contains excerpts. Encounter picked up 113 survivors and was supposed to return them to Batavia, but upon hearing of a “strong enemy force” to the west, returned to Soerabaja instead. Morison, p. 357.

[56]. Womack, p. 126.

[57]. H.J. Hoowij’s illustration of the cruisers passing Kortenaer survivors shows the forward guns of De Ruyter trained to starboard.

[58]. ONI Narrative, p. 75. The radio log of the John D. Ford shows an almost identical message “BT target on port, four points, VA.” Report of John D. Ford (228): “Report of Battle of Java Sea, forwarding of: Extracts from radio log 26 February to 1 March 1942.” Enclosure (B). The origin and intended recipients are unknown. This may have been Perth providing a voice message for Java and other ships in the vicinity.

[59]. Id. “Four points” is a reference to the 32-point compass rose, traditionally used in maritime navigation for both true and relative bearings. One point is equal to 11.25 degrees, thus four points equals 45 degrees and eight points equals 90 degrees.

[60]. Dull, p. 84; Anthony P. Tully, “Naval Alamo,” found at the US Asiatic Feet Web site: http://www.asiaticfleet.com/.

[61]. Hara, p. 76. It must be pointed out that Hara actually indicates that Takagi ordered Sentai 5 to slow down after it had reversed course and headed north. This does not make sense. Other reports give the impression of a running gun battle that lasted some 20 minutes, and the proposed torpedo firing solutions that the Japanese ultimately used involved a partial stern chase of the Allied cruisers. It was not an ideal solution and further suggests that Takagi had to catch up to the Allied column.

[62]. Winslow, Ghost, p. 123.

[63]. Waller’s report, p. 141. The evidence disagrees as to when the exchange of starshells occurred, whether it was before or after Sentai 5 reversed course. I have decided to go with before the reversal of course, on the logic that it was common practice when making first contact with the enemy at night to try to illuminate them.

[64]. CombinedFleet.com lists the top speed of the Myoko-class cruisers as 34 knots. No specific speed is given for the ABDA cruisers, but Kroese, p. 89, quotes survivors of Kortenaer as saying the cruiser column passed them at “top speed.” Royal Netherlands Navy Warships of World War II lists De Ruyter’s top speed as 32 knots and Java’s top speed as 31 knots. “Top speed” in naval jargon does not necessarily always mean the best possible speed, but given Doorman’s warnings about the slowness of Kortenaer, it is probably safe to assume the Combined Striking Force was running at a speed of at least 30 knots. If Sentai 5 was trying to catch up with the Combined Striking Force, Nachi and Haguro had to be running at or near their top speed, likely 32-34 knots. The Japanese apparently pulled just about even with the Allied column after about 20 minutes. For a 20 minute chase the Japanese could have knocked more than a mile off the range.

[65]. Toland, p. 261; G. Hermon Gill, Australia in the War of 1939-1945, Series Two: Navy; Volume I: Royal Australian Navy 1939-1942, Canberra, Australian War Memorial, 1957, p. 615. Sources do not agree as to which direction the Japanese cruisers turned. I have gone with Dull, p. 85.

[66]. Australian diver and photojournalist Kevin Denlay, who dove the sunken wrecks of De Ruyter and Java as part of an expedition from the southeast Asian wreck diving ship MV Empress in December 2002, found and photographed Java’s portside guns pointing “defiantly upwards.” This position strongly suggests Java continued to train her guns on the Japanese cruisers at her maximum range. Kevin Denlay, “Cruisers for Breakfast,” Sportdiving Magazine, 2003, pp. 17-19.

[67]. It is interesting that this particular message was recorded when so many others were not.

[68]. Marvin Sholar, a US Navy signalman who had been assigned to De Ruyter as a communications liaison, said Doorman “suspected another torpedo attack and … turned away.” Mullin, p. 227. There is no agreement as to the firing solution used by the Japanese. Waller’s report says only that the Japanese ships were “a long way off.” Morison says the torpedoes were fired "when the two columns were almost parallel, 8,000 yards apart.” Gill cites one Japanese report as saying 10,000 meters (10,936 yards), adding that the torpedoes struck “after ten or fifteen minutes, the time estimated for them to reach their targets.” Gill also cites another, “more detailed Japanese report” as stating: "00.53 torpedoes started being fired (Nachi 8, Haguro 4) shooting angle 80 degrees, distance 9.5 kilometers" (10,389 yards). Hara, p. 76, whose figures were later used by Hornfischer, says the solution was shooting angle 60 degrees at 10,000 meters. Though usually reliable, Hara’s figures here must be questioned as would have gotten this information secondhand because he himself was commanding Amatsukaze during the battle. Additionally, 60-degree firing angle, while acceptable, would have had the torpedoes chasing the Allied ships northward somewhat. I find an 80-degree angle more believable, therefore I use the figures cited by Gill – 80 degrees, 9.5 kilometers.

[69]. Takagi may not have had much choice due to his lavish use of his 8-inch ammunition. Hara, p. 78, states that at the end of the battle Nachi had only 70 8-inch shells left. Lacroix and Wells, p. 298, appear to disagree, noting that for the Battle of the Java Sea, Nachi and Haguro, each with a supply of about 1,300 8-inch shells, fired 845 and 774 shells, respectively. Lacroix and Wells go on to say that between 5:47 and 6:50 – the afternoon action – the Japanese fired 1,271 shells, of which only 5 hit, and 4 of those were duds. They attribute the lack of success to the extreme range. According to Hara, p. 78, Takagi was heavily criticized for “his series of blunders,” which included opening gunfire at extreme range and wasting ammunition. One disgusted gunnery officer said of Takagi, “He’s a submariner, and doesn’t know how to use guns.” Lacroix and Wells also point out that of 153 Type 93 torpedoes fired by the Japanese during the Battle of the Java Sea, only three hit – but each of those three hits was fatal.

[70]. The 90-degeee turn suggests that Takagi’s cruisers had pulled even with De Ruyter by this time.

[71]. Pretty much all histories, either in narrative or in map form, have the Combined Striking Force at this point remaining in column and executing a “column turn” or what Morison calls a “column movement” – basically a follow-the-leader move in which each ship turns at the same point on the water and into the same direction as the ship in front of it. The ONI’s diagram illustrating this portion of the action follows this scenario. However, what actually seems to have happened is a “line turn,” a “simultaneous turn,” because ideally all the ships are supposed to turn simultaneously; an “echelon turn” as Jonathan Parshall and Anthony P. Tully, Shattered Sword: The Untold Story of the Battle of Midway, Washington, DC; Potomac, 2005, p. 346 describe such a maneuver, or what Morison simply calls a “turn” – each ship turning in the same direction simultaneously or as close to simultaneously as they can manage. Morison, p. 347, n. 11. Obviously, this represents a major change in the accepted narrative, but it fits both the known facts and tactical doctrine. Such a major change obviously will require justification. For simplicity’s sake, I will give the general reasons here; the details will follow in this and subsequent endnotes.

1. The standard response to a torpedo attack was an immediate turn to comb the torpedoes. Doorman had already used such a tactic late that afternoon during the final Japanese torpedo attack ending the day action.
2. The Combined Striking Force had also mistakenly executed a line turn when Exeter sheared out of column after being hit by Haguro. Doorman had to circle his confused ships to reform the column. So the line turn and the reform were something the crews were used to and expected.
3. A ship captain is very unlikely to maintain a course knowing torpedoes are inbound for the sole purpose of staying in column.
4. Survivors of Houston reported that De Ruyter was ahead of her and “slightly to the left,” when the Dutch cruiser was hit. Suggests a slight right echelon formation.
5. The firm belief on the part of the Houston survivors that they were immediately behind De Ruyter suggests they had an unobstructed view of the flagship when she was hit. This means Perth was not directly between Houston and De Ruyter, which means she was somewhere to port. Again, suggests a right echelon formation.
6. A line turn explains the heretofore unexplained starboard turn of De Ruyter to the southeast as a way to reform the column, as Doorman had done at least twice that day.
7. Given that De Ruyter was headed southeast when she was hit, if Perth was in column behind De Ruyter she would not have had to turn northeast to avoid the stricken flagship, as the ONI’s diagram of the incident suggests.
8. If Perth had been in column, her port turn to avoid De Ruyter would have kept her on the other side of the cruiser from the Japanese. But in the line turn scenario, Perth would have been backlit as she turned northeast. Note that Takagi took his cruisers in pursuit of the remaining ABDA cruisers to the northeast for heretofore unexplained reasons.
9. If Houston had been in column, her starboard turn to avoid De Ruyter would have backlit her heading southeast to the Japanese. This would not have happened during a line turn. Again, Takagi took his cruisers northeast, not southeast.
10. Perth and Houston definitely split up, but instead of splitting up to confuse the Japanese they were forced to do so by De Ruyter stopping in their paths. Perth had been ahead of Houston in the original column, but as their near collision later on suggests, ended up significantly behind Houston. Yet there is no mention anywhere of Houston passing Perth.
11. The ONI’s diagram (and many subsequent illustrations) shows Java turning separately from the rest of the column. There is no logical reason for her to turn separately while the other three ships stay in column.
12. Finally, the order for such a maneuver could have been quickly passed back through the column in a similar manner to that used by the Japanese during the collision between Mikuma and Mogami after the Battle of Midway in June 1942.

From a tactical standpoint, a column turn in this situation simply makes no sense. Combing the torpedoes required an immediate turn or risk a torpedo hit, which neither Waller, Rooks nor van Straelen would do for the sole purpose of staying in column. Furthermore, the subsequent movements of De Ruyter, Perth and Houston as reported make much more sense if this was a line turn instead of a column turn. Finally, on at least two occasions during the afternoon action, the Combined Striking Force executed line turns, one accidental, the other to avoid torpedoes.

The accepted scenario of a column turn appears to be based on a misinterpretation of Waller’s report. Waller, p. 141, said he and “followed” the De Ruyter’s turn. Waller goes on to say that Java was struck “whilst the line was halfway round this turn.” Collins, p. 138, said the remaining cruisers “conform[ed].” The ONI Narrative and later historians interpreted that to mean Perth stayed in column, but Waller did not actually say that. His statement that he “followed” De Ruyter likely means he turned after the Dutch cruiser did and remained behind her, but not in column. While “the line was halfway round this turn” can be interpreted as a column turn, this is not necessarily so, because the line turn would also have been sequential as the order was passed back. In the line turn scenario, Perth would have been in a slight echelon to starboard.

[72]. The scenario here is a supposition based on the scenario of the Mogami-Mikuma collision after the Battle of Midway. Parshall and Tully, pp. 345-346; Fuchida Mitsuo and Okumiya Masatake, Midway: The Battle That Doomed Japan, Annapolis, Naval Institute Press, 1955, p. 190. In the foggy early morning hours of June 6, 1941, the Japanese 7th Cruiser Division while proceeding in column sighted a US submarine. Flagship Kumano ordered evasive action in the form of an immediate 45-degree turn to port. She flashed that signal to Suzuya behind her and began turning. Suzuya, in turn, flashed the 45-degree turn order to Mikuma behind her and began her own turn. Mikuma then flashed the order to Mogami behind her and began her turn. Then Mogami began turning.

But in the foggy predawn darkness the move went awry. Kumano’s turn was so wobbly that she nearly collided with Suzuya, who had executed the turn perfectly. Mikuma turned 45 degrees and found herself about to collide with Kumano so she turned another 45 degrees to port. Mogami, last in the column, turned 45 degrees as ordered, but thought the other ships were proceeding without her and may have mistaken Suzuya, ahead of her, for Mikuma. As a result, she did not see that Mikuma had turned too far, had turned onto a collision course until Mogami saw Mikuma’s beam slide in front of her. By then it was too late.

The resulting collision left Mogami’s bow bent forward of her No. 1 turret and limited her speed. Mikuma suffered minor damage in the form of a pierced oil tank, but that was by far the worse injury. Mikuma left an oil slick that US carrier attack planes followed to batter the two cruisers. Mikuma was sunk, and Mogami was so badly damaged that she was, quite literally, never the same ship again.

The Mogami-Mikuma collision helps illustrate the dangers in executing the line turn, even among ships with common signals and experienced crews who had trained together for a very long time. While the Combined Striking Force, as critics claim, may not have trained together very long, they appear to have executed the maneuver perfectly, which should be a testimony to the skill of their crews.

[73]. Winslow, Ghost, p. 124.

[74]. The ONI Narrative and most subsequent illustrations of this action show Java turning separately from the other three cruisers. Why Java would have turned separately has never been explained and on its face does not seem to make sense. It only makes sense if all four cruisers had turned separately.

[75]. Tully, “Naval Alamo; Van Oosten p. 116.

[76]. Mullin, p. 224.

[77]. Hornfischer, p. 92.

[78]. Id.

[79]. Denlay, p. 17, 20.

[80]. “Java,” Royal Netherlands Navy Warships of World War II http://www.netherlandsnavy.nl/ recovered 8/19/08.

[81]. Winslow, Fleet, p. 210.

[82]. The attack had also severely damaged the light cruiser USS Marblehead, forcing her retirement.

[83]. ONI Narrative, p. 76; Tully, “Naval Alamo.” Sholar said De Ruyter was hit “as we were turning back.” Mullin, p. 227. De Ruyter’s turn here is rarely even referenced, and I have not found a reason suggested for this turn. If one accepts the premise of a line turn, then a likely reason becomes apparent: to reform the column.

[84]. This is a supposition that happens to fit the facts and the scenario, but I must admit to finding no hard evidence to support it. Doorman and his staff were well aware of torpedoes in the area and had a rough idea of their track. He would not have knowingly turned De Ruyter into their firing track. Houston had seen torpedoes pass her. De Ruyter may have as well. For Doorman to have turned the cruiser in the very tight timeline provided by the known facts, they likely saw Nachi’s torpedoes pass.

[85]. Mullin, pp. 226-227. Denlay, p. 16, 19, photographed and describes the wreck of De Ruyter with her forward guns pointing to starboard.

[86]. ONI Narrative, p. 76. This information makes it possible to calculate De Ruyter’s heading when she was hit. If the Japanese report (quoted by Gill) of an 80-degree firing angle is used, De Ruyter would have been heading 125 degrees True. If Hara’s 60-degree figure is used, De Ruyter was on a heading of 105 degrees True.

[87]. “The Conquest of Java Island, March 1942,” The Dutch East Indies Campaign 1941-42 http://www.geocities.com/dutcheastindies/java.html recovered 8/19/08.

[88]. Tully, “Naval Alamo.” Denlay, p. 18, photographed De Ruyter’s wheel still showing a starboard turn.

[89]. “The Conquest of Java Island, March 1942,” The Dutch East Indies Campaign 1941-42 http://www.geocities.com/dutcheastindies/java.html recovered 8/19/08.

[90]. Van Oosten, p. 116. It must be noted that his is the only source that I have identified to have this one-minute differential. No one else – not Hara, not the Tabular Records of Movement on CombinedFleet.com – show this differential. This one-minute difference, however, would explain why Doorman thought the torpedoes had passed when they in fact had not, and why Haguro’s torpedoes followed such a similar track to those of Nachi. The possible explanations for these characteristics of Haguro’s attack are much more dubious for a simultaneous launch.

[91]. In the absence of other information pertaining to the spacing between Nachi and Haguro, I have assumed a spacing of one kilometer, which was typical for the Japanese. Haguro would have had to travel one kilometer in the one minute between the time of Nachi’s launch and the time of her own. 1 kilometer = 0.5399568 nautical miles * 60 minutes = 32.397408 knots. Again, per CombinedFleet.com, the top speed of the Myoko-class cruisers was 34 knots.

[92]. Mullin, p. 227.

[93]. “De Ruyter (I) History,” Royal Netherlands Navy Warships of World War II http://www.netherlandsnavy.nl/ recovered 8/19/08.

[94]. Winslow, Ghost, p. 124, Fleet, pp. 209-210.

[95]. David Thomas, The Battle of the Java Sea, New York, Stein and Day, 1969, p. 212; Hoyt, p. 256.

[96]. “The Conquest of Java Island, March 1942,” The Dutch East Indies Campaign 1941-42 http://www.geocities.com/dutcheastindies/java.html recovered 8/19/08.

[97]. Winslow, Ghost, p. 124, Fleet, pp. 209-210. Houston does not seem to have had a major problem avoiding De Ruyter. “Partial Log As Kept By Survivors, USS Houston,” Enclosure (b), The Wartime Cruise of the U.S.S. Houston,” (found at Hyperwar) states that the Houston “had to swerve very sharply else we would have hit her.” But aside from that blurb, none of the Houston survivors mention any particularly violent or sudden maneuvers to avoid De Ruyter. Waller, p. 142, says the Houston headed out to starboard. Schultz, p. 160, quotes a survivor as saying, “The De Ruyter was ahead of us and slightly to our left” when she was hit. Schultz, p. 61 also quotes a survivor as saying, “We saw the tracks of torpedoes astern of us. They just went harmlessly on by.” Both statements put Houston to starboard of De Ruyter and not in column. Winslow says that De Ruyter had changed course to starboard and “[T]he Houston was about to follow when the flagship was hit. ”The ONI Narrative, p. 76, merely says Houston “turned out of column to starboard.” However, this version of events is contradicted by the ONI Narrative’s own chart, which shows that neither Houston nor Perth had any reason for violent maneuvers to avoid De Ruyter because the Dutch cruiser was not in their path. In fact, according to the ONI chart, Houston would have had a more difficult turn to starboard. Notice here that none of these descriptions mentions where Perth was when De Ruyter was hit, even though if they were still in column Perth would have been right ahead of Houston and at least partially obstructing her view. This leads me to conclude that Houston had an unobstructed view of De Ruyter when the Dutch cruiser was hit, while Perth was somewhere to port. The temporary loss of night vision due to the explosions of Java and De Ruyter may have caused Houston to lose track of Perth in the darkness. Combine this with the near-collision later on as Perth was passing Houston and it is easy to see why the survivors of Houston became convinced that they were immediately behind De Ruyter.

[98]. Waller’s report, p. 142. Waller’s exact words are “I just managed to miss her by the use of full helm and one engine stopped.” Collins, p. 138, clarifies by stating Perth “avoided the blazing wreck by the use of full port rudder and one engine.” The assumption seems to be that Perth had to use these maneuvers to avoid plowing into De Ruyter’s stern from the back, but neither Waller nor Collins actually says that. It is also a bit difficult to fathom such an emergency maneuver to avoid a ship which, if they were in a column formation, would have been presenting a narrow stern profile. Additionally, the ONI diagram shows Perth making a sharp turn to the northeast, but no reason for this turn is apparent since, according to the chart, Perth could have just continued to go east. If Perth had already turned southeast to follow De Ruyter, her port turn would have taken her southeast. I can identify no scenario in which Perth, if the cruisers were still in column, would have had to make an emergency turn northeast. Accepting the premise of a line turn presents what, to me at least, is a much more believable scenario. De Ruyter would have presented her stern jutting northwest toward Perth. What seems to have happened is that Perth was in danger of running into De Ruyter’s stern from the side and, given her high rate of speed, possibly shearing it off. This required Perth to make a very difficult turn to port, as Collins describes, which was northeast, as the ONI chart shows. It must be pointed out also that none of Houston’s survivors mention seeing Perth having difficulty avoiding De Ruyter.

[99]. Hornfischer, p. 92.

[100]. Hara, p. 76.

[101]. Hara, p. 76.

[102]. This scenario is based on the circumstantial evidence of Takagi’s northeast turn, but it is also deductive. If one accepts the premise of a line turn, Perth turning northeast to avoid De Ruyter would have been between the Dutch cruiser and the Japanese, and would have been silhouetted by the fires of the blazing wreck. Houston turning to the southeast would not have been backlit. If the Combined Striking Force had remained in column, Perth following her port turn would have been on the other side of De Ruyter from the Japanese and would not have been seen. Houston, on the other hand, would have been silhouetted as she passed the blazing De Ruyter to starboard. Given that De Ruyter was on a southeast heading when she was hit, it would have made more sense for the Japanese to look southeast for Perth and Houston, not northeast, unless they actually saw Perth turn northeast, which would not have happened if the Allied cruisers were in column.

[103]. Tully, “Naval Alamo;” Bosscher, p. 291. Schultz, p. 161, has the message as “Do not stand by for survivors. Proceed to Batavia.”

[104]. Tully, “Naval Alamo.”

[105]. Waller, p. 142.

[106]. Morison, p. 357. Morison states that Perth and Houston had “separated, hoping to shake off the tracking enemy planes.” But this is incorrect. By this time there were no Japanese planes aloft tracking the cruisers. Additionally, the cruisers were actually forced to separate by their maneuvers, especially those of Perth, to avoid De Ruyter.

[107]. Winslow, Ghost, p. 124-125; Schultz, p. 161, quotes a Houston survivor as saying “We just barely missed her stern by about two feet.” This incident is usually associated with Perth’s maneuvers to avoid De Ruyter, as both Winslow and Schultz do, in part because Captain Rooks was trying to avoid torpedoes, but under either a line or column scenario this incident must have happened much later. If the ships had been in column when De Ruyter was hit, Perth would have been ahead of Houston. She did turn to port to avoid De Ruyter. The near-collision requires Perth to have been behind and to starboard of Houston. The survivors of Houston make no mention of having passed Perth; they believed Perth was behind them all along, and this incident undoubtedly contributed to that belief. In a column formation, Houston could have passed Perth while both were passing De Ruyter, but that would have had Houston avoiding De Ruyter, not Perth and still left Perth to port of Houston. Under the line turn scenario I advocate here, Perth turning to the northeast to avoid De Ruyter would have left her well behind Houston. The narrow right echelon formation the ships were in would have meant Houston would have been more focused on De Ruyter than Perth. Again, none of the survivors of Houston mention passing Perth or even seeing her struggle to avoid De Ruyter. It should be mentioned here that this near-collision by itself shows Houston had completely lost track of Perth, even though the Australian cruiser was close by to starboard in the darkness. Houston most definitely could have lost track and likely did lose track of Perth somewhere to port as the cruisers tried to comb the Japanese torpedoes before the hit on De Ruyter. Finally, at this point in the battle, there were no Japanese torpedoes in the water. False sightings are not uncommon in war.

[108]. The Last Stand of the USS Houston, DVD, Jason Eisenberg, released May 31, 2006.

[109]. In his report, Waller wrote on his decision to withdraw:

I now had under my orders one undamaged 6-inch cruiser, one 8-inch cruiser with very little ammunition and no guns aft. I had no destroyers. The force was subjected throughout the day and night operations to the most superbly organized air reconnaissance. I was opposed by six cruisers, one of them possibly sunk, and twelve destroyers. By means of their air reconnaissance they had already played cat and mouse with the main striking force and I saw no prospect of getting at the enemy (their movements had not reached me since dark, and even then the several reports at the same time all gave different courses) . It was fairly certain that the enemy had at least one submarine operating directly with him, and he had ample destroyers to interpose between the convoy and my approach – well advertised as I knew it would be. I had therefore no hesitation in withdrawing what remained of the Striking Force and ordering them to the pre-arranged rendezvous after night action – Tanjong Priok. Waller, p. 122.

Waller was probably anticipating the rebuke he would receive from Dutch Admiral Helfrich. His comments, and editorial responses from Gill are as follows:

Strictly speaking the return of Perth and Houston was against my order 2055/26 – “You must continue attacks till enemy is destroyed." This signal was intended to make it quite clear that I wanted the Combined Striking Force to continue action whatever the cost, and till the bitter end. Perth did receive this signal. Both cruisers were undamaged [Houston’s after triple turret was out of action] and it was not right to say in anticipation “It is no use to continue action”, considering the damage inflicted upon the enemy cruisers, which in my opinion must have been severe. [Actually the enemy cruisers were all in battle trim.] However, it is possible that other facts had to be considered, such as shortage of fuel or ammunition. [Houston, as stated above, had very little ammunition remaining.] The decision of the captain of Perth is even more regrettable as, after all, both cruisers did meet their end. Probably on the night of 27th-28th February they would have sold their lives at greater cost to the enemy. Gill, p. 616.

Gill added his own comments, including an invocation of Thermopylae:

In his desire for “the Combined Striking Force to continue action whatever the cost, and till the bitter end”, Helfrich disregarded a major point in warfare: “When is it the right time to disengage?” On numerous occasions in the history of battles, he who found the right answer to that question has been rewarded with victory – a prize that has seldom been given in recognition of military suicide. Here were no conditions warranting a Thermopylae, with commensurate rewards for the sacrifice. Had none but military considerations governed the use of the Allied naval forces in the Java campaign, the time for their disengagement and withdrawal was reached long before Waller took his absolutely correct action in disengaging and withdrawing the remnant under his command. In that action he did his duty to the Allied cause, which would have been much better served by his saving the two ships and their trained crews for future use. In the event they were lost twenty-four hours later; but even so, most probably at greater cost to the enemy than would have been had Waller decided in favor of an unrealistic gesture on the night of 27th-28th February. (Id.)

As the fall of Java became more and more certain, Helfrich’s actions bordered more and more on the irrational. On top of these sentiments, after the Battle of the Java Sea Helfrich ordered Perth and Houston to go through the Soenda Strait not to withdraw, but to go to Tjilatjap to continue fighting the Japanese after they had landed on Java. Neither Waller nor Rooks thought Helfrich was serious, that it was preparatory to withdrawing them to Australia. It was in transiting the Soenda Strait where Perth and Houston ran into an invasion convoy and sunk in the epic stand on the night of February 28-March 1 1942.

Additionally, Helfrich, on at best questionable authority, had ordered the US seaplane tender Langley, formerly the US Navy’s first aircraft carrier, and the freighter Seawitch to ferry P-40 Warhawk fighters to Tjilatjap. The Langley was carrying 32 P-40’s with pilots, but while the fighters were desperately needed, Tjilatjap had no airfield and no way to get the P-40’s to one. Langley was sunk on February 27 by Japanese aircraft operating out of Celebes. Seawitch made it to Tjilatjap, but her aircraft were crated and required assembly. With no time left before the Japanese invasion, the crates were dumped into Tjilatjap’s harbor. There is some belief the Japanese may have recovered and assembled them.

By such actions, Helfrich indicated a willingness to fight to the last Dutch – and American. And British. And Australian. In fairness to him, Helfrich was a European native of Java and was not alone in his incredulousness at the situation in which the Dutch found themselves. Additionally, once the British informed Helfrich of their intention to withdraw, Helfrich ordered the Americans to do the same, and thus spared the US officers pangs of conscience at possibly having abandoned the Dutch. Helfrich ultimately fled to Ceylon where the few remaining Dutch ships fought with the British Far Eastern Fleet. He accepted the Japanese surrender on behalf of the Dutch in 1945.

[110]. “Partial log as kept by Survivors (of the USS Houston),” ONI Narrative, Enclosure (a)(9).

[111]. Womack, p. 126.

[112]. Hornfischer at 94.

[113]. Id.

[114]. “De Ruyter (I) History,” Royal Netherlands Navy Warships of World War II http://www.netherlandsnavy.nl/ recovered 8/19/08.