Black May (24 page)

By 1800
^*
on 3 May the new
Specht-Star
rake ran from 56°21'N, 44°35'W (on the German grid AJ 5333) to 54°57, 39°35'W (AK 4449). Boats from this formation reported seeing smoke clouds and starshells; one signaled it had been driven off by a destroyer. The supposition in Berlin was that these boats were in contact again with SC.128. With a note of frustration, if not desperation, BdU signaled: DO NOT HOLD BACK…. SOMETHING CAN AND MUST BE ACHIEVED WITH 31 BOATS.
⁶
Berlin estimated that the convoy was steaming on a course between 20° and 50°. But it was not. While some of its escorts took a course northeastward,
firing starshells to draw off the U-boats, SC.128, alerted to
Specht-Star
s estimated position by Canadian Naval Service Headquarters, Ottawa, which had DFed it, took a jog to the west before resuming a northerly course and then turning east above the north end of the rake. Successfully evading
Specht-Star
, the convoy would arrive at Liverpool on 13 May without mishap. In breaking off the hunt, BdU noted, “Most of the boats are short of fuel, and it is pointless for them to run about after the convoy.”
⁷

At the same time BdU formed
Specht-Star
it also augmented
Gruppe Amsel,
to the south, and formed it into four subgroups, I, II, III, and IV, of five U-boats each, except for I, which had six.
Amsel
now ran, with gaps between the subgroups, from 51°51'N, 49°05'W (AJ 7933) to 44°15'N, 39°35'W (BC 9646). In a revealing comment about BdU’s awareness of the Allies’ shore-based HF/DF capability, the Berlin war diary observed: “This new type of disposition should avoid the drawbacks that arise when a patrol remains in one place for a long time so that it is D/Fed, sighted, located, etc. by the enemy, who thus finds out its entire extent.”
⁸
The boats at the extreme ends of this segmented line were supplied with dummy F.T. messages with which to create the impression of a larger line “stretching right around the Newfoundland Banks.” That impression was not unlikely to be made, since the OIC Tracking Room was now estimating the number of U-boats at sea to be 128, the highest ever known, representing nearly 60 percent of the Atlantic operational force. When the Allies DFed the boats forming
Amsel,
Dönitz and Godt expected they would discover the gaps and attempt to vector convoys through them. The plan then was to combine the subgroups into a closed line. Before it had a chance to work, however, that plan was overtaken by a new plan, as BdU realized, on 4 May, that the
Amsel
boats would be needed in operations to the north.

At 1602 GST (1402 GMT), Berlin began to reorganize most of the
Specht
and
Star
boats into a new reconnaissance line code-named
Gruppe Fink
(Finch). Ordered to be in place by 1000 GST (0800 GMT) on 5 May, the twenty-seven boats of
Fink
would occupy stations along a line running from west-northwest to east-southeast, or precisely, from 56°45'N, 47°12'W (AJ 2758) to 54°o9'N, 36°55'W (AK
4944)-When formed, the patrol line would stretch 382.6 nautical miles (nm), with an average spacing of 14.7 nm between the boats.
⁹
As these boats were moving into position on the afternoon of 4 May, several (U
—264,
U
—628,
U
—260,
U-270) reported sighting destroyers (Offa or
Oribi,
or both) on southerly courses. Then U
-628
(Kptlt. Heinz Hasenschar) in quadrant AJ 6271 (55°40'N, 42°40'W) at the near center of the
Fink
line reported at 2018 GST the mast tops of a southbound convoy that BdU had been expecting by dead reckoning, that is, by calculation based on a convoy’s course, speed, and elapsed time from a previously determined position.

This was ONS.5 (No. 33), except that BdU mistakenly called it ON.180 (convoy No. 36), which was the convoy that had been trailing ONS.5, but which on 4 May was considerably to the north tracking a WSW course through U-boat quadrants AJ 22 and 23, south of Cape Farewell. BdU was also mistaken in both its dead reckoning and real time calculations, for it expected the convoy reported by U—
628
to cross the
Fink
recco line on 5 May, when, in fact, ONS.5 would reach the center of that line by the late afternoon of the 4th, before
Fink
was fully formed; and if ON.180 had continued to follow ONS.5’s course it would not have reached the line before 6 May. Apparently assuming that convoy No. 33 (ONS.5) had already passed through the
Fink
position, BdU’s dead reckoning error with respect to this convoy may have occurred because it was not aware that during the period 0800 GMT 1 May to 0800 4 May, ONS.5 was practically hove to in contrary weather at speeds no greater than 2.7 to 3.1 knots.
¹⁰

As late as 6 May, when BdU did a wash-up (postaction analysis) on this convoy (
Abschlussbetrachtung Geleitz. 36),
it still identified it as ON.180; but in communications to
Fink
boats during 5/6 May and in the war diary of 26 May it called it the “Hasenschar convoy,” after the Commander of U
-628,
who had been the first to sight ONS.5, at 2005 GST, and to report it, at 2018, on the 4th.
¹¹
The BdU practice of identifying a convoy by its shadower was common. Immediately upon Hasenchar’s report that a convoy was southwest-bound on course 200°, speed 7 knots, BdU ordered up the northernmost subgroups
Amsel I
and
II,
as well as the independently operating U
-258
(Massenhausen),
which had sunk
McKeesport
, and
U-614
(Kptlt. Wolfgang Sträter), which had been temporarily hors de combat with engine problems, to join the
Fink
line. Twenty-seven boats strong on the night of 4 May,
Fink
would eventually claim a total of forty-one boats, the largest concentration ever arrayed across and around a single convoy. Dönitz and Godt reinforced this fact in a signal to the massing boats:
YOU ARE BETTER PLACED THAN YOU EVER WERE BEFORE.
¹²
But BdU worried that owing to fuel depletion several of the boats would not be able to operate much longer than they had.

Of these changing U-boat dispositions the OIC Submarine Tracking Room in London and thus Western Approaches had no direct knowledge until after GC&CS made a break back into naval Enigma at noon on 5 May. To that point, as Commander Winn lamented, “Nothing is known from Special Intelligence of the operations during this period.”
¹³
When, however, GC&CS could read German traffic again, the time lag between interception and decryption was so great— from seventeen hours to twelve days, the norm being four days—the information had no operational value in the battle then joined. It is possible at this date to read the GC&CS decrypts crafted afterwards of the traffic that had passed during the blackout period. Similarly, one can consult the American decrypts of the same traffic that date from later in 1943 when U.S. Navy cryptanalysts acquired raw Enigma intercept material as well as their own “bombes” (decryption machines), hence an independent capacity to make penetrations into the German naval cipher Triton.
¹⁴
But none of that intelligence was available at the time of battle. The principal value of Ultra in the Atlantic struggle had been its strategic disclosure of U-boat positions, and of their operational instructions from BdU. That value was lost on ONS.5. But not everything was lost. Once a close battle was joined, timely and localized intelligence such as that derived from shipborne HF/DF, radar, and asdic was far the more valuable, and that ONS.5’s escorts could collect.

By dusk on 4 May, Sherwood in
Tay
had ample indication that he was in the neighborhood of a large U-boat formation. His FH3 HF/DF was picking up contacts on the port bow, port quarter, starboard beam, and starboard quarter. He was restricted from gaining
accurate fixes, however, by the fact that communications failed between
Tay
and FH3-equipped
Oribi,
resulting in
Tay
obtaining only one cross-cut fix in the next three days, 4 to 6 May. If Sherwood needed any confirmation from afar that he was surrounded, it came from the Admiralty, which signaled him at 1920 about the existence of heavy and continuous W/T traffic in his vicinity on 12215 and 10525 kilocycles.
¹⁵
Two sweeps by the Support Group destroyers
Offa
and
Oribi
failed to locate any of the sending boats. The convoy was still east of 47°W and north of 4o°N, beyond which boundaries, west and south, the new Canadian North West Atlantic (CinC, CNA) Command governed all surface and air anti-submarine escorts, as decided by the Washington Convoy Conference of 1–13 March 1943. (That conference, attended by senior British, Canadian, and American naval and air representatives, also decided, among other things: that the British and Canadians would share command of the northern Atlantic convoy lanes, while the United States would concentrate her forces in the central Atlantic, including the routes of the tanker convoys between the West Indies and Britain; and that [at last] 255 VLR aircraft would be delivered to the airfields on both shores of the Atlantic by July.) In the longitudes where ONS.5 sailed during the critical days of 4–6 May her escorts still remained under British operational control, although the Admiralty’s counterpart OIC in Ottawa, employing a high-power, low-frequency transmitter near Halifax, communicated HF/DF-derived U-boat position estimates to convoys, such as SC.128, eastward as far as 30"W.
¹⁶

Whereas SC.128 had been rerouted to evade the DFed
Specht
line, the suggestion has been made that ONS.5 was not similarly vectored around DFed boats before the evening of 4 May because of the escorts’ low fuel levels, and their need to continue on the shortest possible route to port.
¹⁷
But with so many boats in movement across nearly 400 miles of ocean, there is a question if either Liverpool or Ottawa knew what possibly would have been an evasive route. The shore-based HF/DF accuracy was reported by the Admiralty to be no better than within 120 miles. Even the Admiralty message to
Tay
at 1920 on 4 May expressed itself as being uncertain if it was ONS.5 or SC.128 that was being shadowed, so “very poor” were D/F conditions. Convoy SC.128 at the time
was approximately thirty miles north of the
Fink
line, traversing squares AJ 28–29–34 on a course northeast by east.

Escort Group
By
now readied itself to run the gantlet. With 30 merchant ships present in ten columns, five cables (3,040 feet) apart, on course 202°, speed seven knots, in very clear weather, wind Force 2—a light breeze, four to six miles per hour—and a slight sea with low, long swell, Sherwood placed his night field as follows:
Sunflower
on the port bow,
Snowflake
on the port beam,
Tay
on the port quarter,
Vidette
on the starboard bow,
Loosestrife
on the starboard beam, and
Northern Spray
on the starboard quarter—although the rescue trawler, which was astern, was delayed in taking up her station. Destroyers
Offa
and
Oribi
provided forward cover on the starboard and port bows, respectively, at five miles distance.
Pink
was leeward at 56°32'N, 40°50'W on course 235° with four stragglers, speed 5 knots; Sherwood recommended that she be separately routed, as was done.
¹⁸

No doubt the shore commands that watched this confrontation of forces unfold, whether at Liverpool or London, Halifax or Ottawa, where enemy dispositions could only be guessed at on their wall charts and plotting tables, held their breath as the volume of HF/DF contacts mounted. At sea the incoming Morse traffic was just as ominous. Said Captain J. A. McCoy, SO, EG3 Support Group, on
Offa
: “During all this time enemy W/T transmissions had become more and more frequent….”
¹⁹
There was no doubt on his bridge that a multitude of foes was thickening around them.

One of the
Specht
boats proceeding to form
Fink
never made it to the party. It was taken out of the fight earlier in the day to the north-northeast of
Fink
, or about thirty miles astern of the convoy, by one of two Royal Canadian Air Force Canso A’s (as the Canadians called the PBY-5A amphibious Catalinas) that came from Gander, Newfoundland, to give ONS.5 its first real air cover in two days, although neither aircraft met the convoy as such. The Air Gap had greatly narrowed during April and early May, with the result that on no day during its transit of the gap did ONS.5 miss contact of some kind with aircraft: even on 3 May a U.S. Army Air Force (USAAF) B-17 Flying Fortress
from Gander rendezvoused with the convoy at 1538, though, at the boundary of its Prudent Level of Endurance (PLE), it could only remain with the convoy for six minutes.
²⁰
Nonetheless, the flyover must have caused the
Specht-Star
boats to keep their heads down.