CHAPTER XI

Clearing the Aleutians

As soon as the decision to occupy Amchitka was taken in December 1942, preliminary planning to drive the Japanese out of the Aleutians was set in motion. Initially Kiska, the nearer, more strongly fortified of the enemy-held islands, and offering a more satisfactory harbor and better airfield sites, was the objective of the counterassault. As a starting point, General DeWitt proposed to organize and train a task force built around one infantry division and totaling 25,000 men. For commander and assistant commander he recommended Maj. Gen. Charles H. Corlett and Brig. Gen. Eugene M. Landrum, both of whom had participated in joint amphibious exercises and were familiar with conditions in the Aleutians. Although Admiral Nimitz, estimating Japanese strength on Kiska at 10,000 men, suggested that two divisions might be required, the War Department concurred in the outline plan presented by General DeWitt. In place of the 35th Division, originally recommended, the War Department proposed, and General DeWitt agreed, to employ the 7th Division, since it was in a better state of training and readiness, was scheduled for early "demotorization" and could be brought up to full strength more readily, was stationed near Fort Ord, where the amphibious training was to be conducted, and was more ably led and staffed.1  

A joint Army-Navy planning staff was set up at San Diego under Rear Adm. Francis W. Rockwell, commander of the Amphibious Force, North Pacific, who was designated to command the assault force for the actual operation. Maj. Gen. Albert E. Brown, commanding general of the 7th Division, was named commander of the landing force. While Admiral Rockwell and a group of officers from the Western Defense Command were making plans, with the help of several Alaskan experts from General Buckner's headquarters, General Brown was leading his troops through the

[277]


amphibious training course directed by Maj. Gen. Holland M. Smith, USMC. By the beginning of February the forces training at Fort Ord for the descent on Kiska included, in addition to the 7th Division, the 184th Infantry Regiment, the 78th Coast Artillery (AA) (less one battalion), and the 2d Battalion, 501st Coast Artillery (AA).2  

Meanwhile, in the Aleutians the Eleventh Air Force had been sending its planes over Kiska whenever the weather permitted, which in December and January was not often. Fog and foul weather held the planes to the ground during most of January, and the few missions that were flown proved more costly to the Eleventh Air Force than damaging to the enemy. Only about 10 tons of bombs were dropped -the lightest since the beginning of the air assault- and at least ten planes were lost, none of them by enemy action. Partly because the weather improved, and partly because P-38 and P-40 fighter-bombers were now based on Amchitka, February and March were much better months. In February Army planes attacked Kiska on nine separate days, flying twenty-four missions (not including twenty weather and reconnaissance missions), and dropping about 150 tons of bombs. The attacks continued with equal vigor and intensity during March.3  There was, to be sure, no comparison between the air assault on Kiska and the huge raids taking place against German-occupied Europe, but it is to be noted that in the South Pacific the Allied air forces loosed 197 tons of bombs on Rabaul during the month of December.

On 26 March a solid opportunity came to the Eleventh Air Force to strike a major blow, when a naval task group cruising off the Komandorski Islands under Rear Adm. Charles H. McMorris intercepted a strong Japanese force that was attempting to run reinforcements into Kiska and Attu. But the opportunity was lost. When Admiral McMorris' report of contact reached Adak, the bombers loaded with antipersonnel bombs were poised for an attack on Kiska. Although General Butler estimated that it would take at least an hour or so to unload the light bombs and replace them with heavy, armor-piercing ones, it seemed logical to accept the delay and make the change.4  Admiral Kinkaid therefore sent a message to McMorris suggesting that he fight a retiring action to the eastward in order to get under cover of

[278]


the bombers, but McMorris at that point was completely cut off by an enemy force twice the strength of his own. Furthermore, the shift of bomb loads cook much longer than anticipated, and by the time the planes were ready a snow storm had closed in the field. When they finally took to the air, they were unable to reach the scene of action before the Japanese had retreated beyond range. Although Admiral McMorris succeeded in thwarting the enemy attempt at reinforcement, the support of Army bombers might have enabled him to turn the engagement, brilliant as it was, into an unmistakable disaster for the Japanese.5  

Attu Retaken

By this time the plans and preparations in motion on the west coast had been given a new objective. Realizing that not enough shipping would be available for the Kiska operation, Admiral Kinkaid had recommended early in March that Attu be substituted as the target, for, in comparison with Kiska, Attu appeared to be weakly defended. Estimates based on air photographs placed the Japanese strength on Attu at only 500 men, of which three rifle companies constituted the effective tactical strength, the remainder being antiaircraft and labor troops. Instead of the reinforced division called for in the Kiska plans, one infantry regiment plus the 7th Division's mountain artillery was considered by Admiral Kinkaid and General Buckner as probably sufficient for the capture of Attu. Only four attack transports (APA's) and two or three cargo ships (AKA's) would be required.6  Also in capturing Attu and with an airfield in operation there, American forces would be astride the Japanese line of communications between the home islands and Kiska. The latter, cut off from supply and reinforcements, would "wither on the vine." On 10 March Admiral King notified Admiral Nimitz and General DeWitt that the joint Chiefs of Staff had approved the projected change of plan, provided the operation could be carried out with only those means that Admiral Nimitz could spare and those already on hand for the assault on Kiska, which was now deferred. The approval of the joint Chiefs, Admiral King made clear, extended only to planning and training; it was not to be considered as a directive to execute the operation.7  This would hinge upon the outcome of the discussions on Pacific strategy

[279]


about to begin in Washington. When the Joint Chiefs of Staff on 21 March agreed to postpone any major offensive in the South Pacific, the way was cleared for the reduction of Attu. Information received by the Navy that the Japanese were establishing an airfield on Attu offered an additional reason for putting the plans into execution. Therefore, on 22 March, General Marshall and Admiral King decided that the operation should proceed "as soon as practicable.8  The shift of objective did not upset the training program or make an appreciable difference in the preliminary planning activity of Admiral Rockwell's joint staff in San Diego. A new estimate of the situation, including a study of the shipping that would be available and the forces required, was necessary. This was prepared by the joint planning staff; and, after the decision to go ahead with the operation was reached, General Brown's staff, in co-operation with Admiral Rockwell's joint planners, began drawing up the detailed tactical plans.9  

Seldom has an operation been planned with less knowledge of the conditions the troops would have to face. From Cape Wrangell in the west to Chirikof Point in the east, the fog-surrounded island of Attu is about forty statute miles in length. Its greatest width is about twenty miles. At the head of the deep coastal indentations lie narrow beaches, from which small, snow-fed streams lead back into the jumbled, barren mountain-mass of the interior, a desolate region of twisted, precipitous crags, whose snowcapped peaks mount upwards to heights of two and three thousand feet. The valley floors are carpeted with tundra: the black muck, covered with a dense growth of lichens and moss, which is characteristic of the far North. On the northern mainland, in Alaska and northern Canada, the tundra is frozen solid during most of the year; but in the outermost Aleutians the Japan Current has a moderating effect on temperatures and much of the time the tundra is barely firm enough for a man to cross it on foot. The same Japan Current accounts for the pea-soup fogs, the constant pervading wetness, and the frequent storms that help to make the outer Aleutians so in-

[280]


hospitable. To the soldiers who had to fight not only the Japanese but the weather and terrain of the island, it must have seemed that the Creator of the universe was an unskilled apprentice when He brought Attu into existence.

The eastern end of Attu, indented by four bays, is roughly shaped in the form of a trefoil: the northern lobe lies between Holtz Bay and Sarana Bay; the elongated midsection, terminating in Chirikof Point, is shaped by Sarana and Massacre Bays; and the southern lobe lies between Massacre Bay and Temnac Cove. (Map III) From Holtz Bay in the north and Temnac Cove in the south, steep-walled valleys run back in a generally westward direction until they disappear in the mountainous maze of the interior. The Massacre Bay valley, about a mile and a quarter wide at the beach, is soon divided into two by a hogback, which, although rather steep on the sides, slopes gradually along its length to an elevation of about 550 feet at the upper end. At Holtz Bay, likewise, a ridge divides the valley into two; but here the central ridge projects into the bay for a distance of nearly a mile, and its highest, steepest sides face the water. About a mile and a half up the west arm of the valley a low pass crosses this ridge into the eastern Holtz Bay valley, from which, at this point, over a 600-foot saddle, it is possible to cross into the head of West Massacre Valley. A slightly lower saddle separates the head of East Massacre Valley from the valley leading out of Sarana Bay.10

Only the bare details of the topography were known to those planning the assault. The only available map of Attu was a Coast and Geodetic Survey chart showing the terrain back to approximately one thousand yards from the shore line, and warning all shipping not to approach closer than two and one-half to three miles. Very little was known about the harbors. Oblique aerial photographs filled in a few gaps, but, because of the prevailing fog, the coverage was far from complete. A terrain model was constructed of the eastern portion of the island, east of a line running from Temnac Cove to the ridge north and west of Holtz Bay, but the model did not clearly delineate the key passes or the areas behind Henderson Ridge (the southwestern wall of Massacre Valley) and in the interior, west of Holtz Bay.11  The American planning staff had only scant information concerning the Japanese defenses. During late fall and early winter the Attu garrison had

[281]


MAP III

THE CAPTURE OF ATTU 7TH INFANTRY DIVISION - 11-30 May 1943

MAP III - THE CAPTURE OF ATTU 7TH INFANTRY DIVISION - 11-30 May 1943

For a larger image click here


been gradually reinforced. A redeployment of naval forces ordered by Admiral Kinkaid shortly after he took command and the subsequent battle off the Komandorskis put an end to the process, but in the meantime the Japanese strength had been increased to a total of approximately 2,400 men. The nucleus of combat troops included about one and a half battalions of infantry, three antiaircraft batteries armed with 75-mm. dual-purpose guns and lighter weapons, and two platoons of a mountain gun battery armed with 75-mm. pack howitzers. In addition to medical and other service detachments there were several engineer units whose primary mission was to construct an airfield at the head of the East Arm of Holtz Bay. The whole force was commanded by Col. Yasuyo Yamazaki, with headquarters at Chichagof Harbor, a small bay midway between Holtz and Sarana Bays. The bulk of the garrison was concentrated in the vicinity of Holtz Bay and around Chichagof Harbor, where the strongest positions had been installed. One of the antiaircraft batteries, consisting of four guns, commanded the West Arm of Holtz Bay; another was placed at the head of the East Arm of the bay; and the third was part of the Chichagof Harbor defenses. The pass between Holtz Bay and Massacre Valley was guarded by the mountain artillery, one platoon of which was in position to enfilade Massacre Valley itself. Along the ridges flanking Massacre Valley and overlooking Sarana Bay the Japanese had built machine gun and mortar positions.12  The plans being developed in California took note of the fact that Holtz Bay and Chichagof Harbor were the most heavily defended of the possible landing places. Reconnaissance planes had noted signs of Japanese activity in the vicinity of Temnac Cove, Sarana Bay, and at the head of Massacre Valley, but it was almost impossible to spot the cleverly concealed emplacements along the ridges. Additional details kept coming to light with the result that the original estimate of Japanese strength was progressively raised. By the time General Brown and his staff had completed the operational planning, it was estimated that the enemy garrison amounted to something between 1,600 and 1,800 men, of whom one battalion or its equivalent was composed of infantry and troops available for infantry service. Aerial photographs received from the planning staffs in early April indicated that a number of Japanese positions existed in the lower part of Massacre Valley commanding the beaches and the bay, but, because there was no sign that these

[282]


positions were occupied, the assumption was that they had been built and abandoned the year before.13  

As soon as it was clear that the Japanese garrison exceeded the first estimate of 500 men, General Brown's landing force was increased by an additional battalion combat team. Thus, for the initial attack, the following troops would be available: one regimental combat team built around the 17th Infantry and a field artillery battalion; one battalion combat team from the 32d Infantry and including a battery of field artillery; the 7th Division Reconnaissance Troop; one battalion of antiaircraft artillery; and one battalion of combat engineers. The remainder of the 32d Infantry, with reinforcements similar to those of the 117th Infantry combat team, was to be held at Adak as a floating reserve and was expected to be available at Attu on D plus 1. The total strength of the assault force and floating reserve amounted to approximately a 11,000 men.14  Admiral Kinkaid, as commander, North Pacific, was in command of the entire operation. Under his direct command were the shore-based air group, the naval escort, cover, supply, and service groups, the floating Army reserve, and a force consisting of the 4th Infantry and one engineer regiment which, after Attu had been taken, was to occupy Shemya Island and construct an airfield there. Under the direct command of Admiral Rockwell was the assault force, which consisted of the naval air and fire support group, the transport group, a mine sweeper group, and the landing force under General Brown, who was to assume tactical command ashore from the time of landing.15  

The lack of information on topography and offshore hazards made it necessary to prepare several optional plans. By the time the main assault force sailed from San Francisco on 24 April five different plans of operation had been worked out. Under Plan A the major landing was to take place in Massacre Bay and a secondary one was to be made at a small beach (Beach Red) 600 yards west of the entrance to Holtz Bay. Under Plan B the major assault was to be launched from Sarana Bay. Plan C was based on landing the entire force in Massacre Bay. In Plan D as in Plans A and B, two landings were to take place: the major one in the West Arm of Holtz Bay and the other at Beach Red. Plan E provided for three landings: one at either

[283]


Beach Red or the West Arm of Holtz Bay, another in Massacre Bay, and a third in Sarana Bay.16  Final decision as to which plan to adopt was postponed until the arrival of the force at its rendezvous at Cold Bay, where it was hoped more reliable information as to navigable waters could be obtained from Aleutian pilots. Admiral Rockwell was inclined to view Plans B and D with disfavor, and General Brown preferred not to depend on a single effort, as in plan C.17  At Cold Bay, a revised Plan E was, after considerable discussion and study, adopted as the plan of attack. Sarana Bay was ruled out entirely. The major landing was to take place at Massacre Bay, and a landing on the north side of the island was to be made wherever it proved most feasible by a reconnaissance on the morning of D-day. The main force, landing at Massacre Bay, was to "advance rapidly" up the valley, seize the passes leading to Holtz and Sarana Bays, and then move into the Holtz Bay area where it was to join the northern force in destroying the enemy in that vicinity. As soon as this had been done, the main force was to advance against Chichagof Harbor, while the northern force secured the valley running west from Holtz Bay.18  The assumption apparently continued to be that not more than three days would be required to take the island.19  

Delayed twenty-four hours because of weather; the attack force headed out of Cold Bay on 4 May and turned westward through chill rain and a stormy sea toward Attu. D-day was set for 8 May. As the force drew near the run-in point 115 miles off the north shore of Attu, the weather grew even worse. Admiral Kinkaid ordered Admiral Rockwell to postpone the landing a day. While Admiral Rockwell took his battleships off to the west on the strength of a rumor that a strong Japanese force was approaching from that direction, the transports and a destroyer screen circled eastward in the dense fog, rain, and rough seas. With the weather continuing foul and reconnaissance planes reporting that a heavy surf was running on the landing beaches, Admiral Kinkaid again postponed D-day. Finally, as there

[284]


seemed to be no prospect of the weather clearing, he ordered the attack to proceed, on 11 May. In the midst of the fog the battleships made rendezvous with the transports on the evening of 10 May, and the force split into two groups for the approach.20  General Brown, who had his headquarters on board the transport Zeilin, accompanied the group heading for Massacre Bay. Admiral Rockwell, in Pennsylvania, remained off the northern coast.21  

The weather had helped to frustrate plans of the Eleventh Air Force for softening up Attu before the assault. The Army had concentrated about two dozen of its most efficient fighter-bombers on Amchitka for preinvasion bombings of the island, and during the ten days preceding the landings Army planes dropped 95 tons of bombs on Attu. But the foul weather that forced the postponement of the landings for four days stopped all attack missions against Attu during the same period, and also most of the more elaborate air support measures planned for D-day.22  

The assault opened according to plan, quietly, like a commando raid, when the 7th Scout Company paddled ashore from submarines in the predawn darkness on a small beach (Beach Scarlet) about four miles northwest of Beach Red, on the north shore of the island. This unit, the 7th Scout Company, was part of a Provisional Battalion commanded by Capt. William H. Willoughby, the remainder of which consisted of the 7th Division's Reconnaissance Troop (minus one platoon). The Reconnaissance Troop, on board the destroyer Kane, was scheduled to follow the Scout Company ashore immediately, but a blanket of fog had again descended on the entire eastern end of the island and the Kane lost its bearing. As a result, the Reconnaissance Troop did not land until nearly noon. By then, the Scout Company had moved well up a steep valley that led south from the beach. At the head of the valley was a pass which gave access to one of the valleys leading back from Holtz Bay, and from which it was hoped the Scout

[285]


Photo: ATTU LANDINGS. Massacre Bay, as the 4th Infantry moved inland.

ATTU LANDINGS. Massacre Bay, as the 4th Infantry moved inland (top). The west arm of Holtz Bay viewed from the ridge over which the troops advanced (bottom). Note crashed Japanese Zero.

Photo: ATTU LANDINGS. The west arm of Holtz Bay viewed from the ridge over which the troops advanced. Note crashed Japanese Zero.

[286]


Company could attack the enemy in the rear. Meanwhile a reconnaissance party of Alaskan Scouts and Company A, 17th Infantry, had groped its way through the pea-soup fog to a landing on Beach Red. Its mission was to explore the feasibility of using this beach to land the entire northern force, a combat team (BCT 17-1) built around the 1st Battalion, 17th Infantry. Two obstacles presented themselves: a rock-studded approach that prevented more than two or three boats from unloading at one time; and a steep escarpment that began about 75 yards from the water's edge and rose to a height of 200 or 250 feet above the beach. From the top of this bluff, a fairly level, but broken, tableland stretched south along the coast to the heights overlooking Holtz Bay. The Navy beachmaster and Col. Frank L. Cullin, commanding officer of the 32d Infantry who had gone ashore with the reconnaissance party, reported that a landing on Beach Red was feasible, and at 1230 Lt. Col. Albert E. Hartl, commanding officer of BCT 17-1, requested General Brown's permission to land the rest of his troops. On board the Zeilin off Massacre Bay, General Brown had been waiting impatiently for the fog to lift enough to permit the main landings to take place. The first assault waves had been on the water since shortly after 0800, awaiting better visibility and a signal to proceed, while H-hour was twice postponed. Finally, at about the same time that he received Colonel Hartl's request, General Brown received a message from Admiral Rockwell advising him that, since the weather now promised to improve, the boats should be sent off as soon as they could feel their way into Massacre Bay. When General Brown was assured that the landing craft could return to the transports for a second trip, he recommended that the Massacre Bay assault begin at 1530 and that Colonel Hartl land his force on Beach Red as soon as he was ready. At 1615, Company B, 17th Infantry, set foot on Beach Red. Minutes later, advance elements of the 2d and 3d Battalions, 17th Infantry, landed on Beaches Blue and Yellow, in Massacre Bay.23

No enemy opposition was encountered at any of the landing beaches. The fog, which hampered the landings, likewise concealed them from the enemy. For several weeks the Japanese had known that an attack on their Aleutian outposts was in the offing, although until the end of April they thought Kiska would be the target. Their first intimation of the American approach to Attu came at 0200 on 11 May, when planes from the carrier

[287]


Nassau combined a bombing and strafing run over Chichagof Harbor with the dropping of leaflets demanding surrender. At 1000, Colonel Yamazaki was informed of American shipping offshore and he ordered combat positions strengthened. Shortly thereafter, battleships Pennsylvania and Idaho engaged in a radar-controlled bombardment of the Chichagof area. The enemy responded by strengthening the defensive positions that guarded the passes leading out of Massacre Valley.24  

By 2130, five hours after the main landings commenced, a total of 3,500 men had gone ashore; 400 at Beach Scarlet, 1,100 at Beach Red, and 2,000 at Massacre Bay. The Northern Landing Force, BCT 17-1, had made contact with the enemy at about 1800, when a patrol party that was moving along the beach at the foot of the escarpment encountered four unsuspecting Japanese about a mile southwest of Goltsov Point. Two of the Japanese were killed; the other two escaped. Soon afterward the beach patrol came under the fire of the dual-purpose guns at the head of Holtz Bay, and its advance slowed. The main body of BCT 17-1, on the tableland above the escarpment, had continued on apparently undetected. Its objective was Hill X, an 800-foot camel back about two miles south of Beach Red, which dominated the Japanese positions at the head of Holtz Bay. By 2230 the gathering darkness merged with the thick fog to disguise the lay of the land completely. BCT 17-1 dug in for the night, not quite sure where it was, but hoping that the hill on which its outpost positions were placed was Hill X.25  The Provisional Battalion, which had landed on Beach Scarlet, had been climbing most of the day up a steep watercourse. By midafternoon the advance unit, the 7th Scout Company, had reached an elevation of nearly 2,500 feet, at what appeared to be the summit of the pass. But from here on the only maps the men had were blank. Rather than risk getting thoroughly lost in the uncharted jumble of peaks, ridges, and canyons that lay beyond, Captain Willoughby ordered his men to bivouac for the night.26  

The Southern Landing Force, advancing in Massacre Valley, had come under enemy fire shortly after 1800. BCT 17-2, which was moving up along

[288]


the right side of the hogback and along the floor of the valley to the east, had advanced approximately 2,500 yards when it was stopped by rifle and machine gun fire coming from the high ridge, later named Gilbert Ridge, that formed the east rim of the Massacre Valley. After being pinned down for about forty-five minutes, the battalion began working forward, although it was still under scattered rifle fire. Immediately the enemy fire became more intense, mortars and light artillery joined in, and BCT 17-2 was again stopped. Unable to move in any direction, at 2100 the battalion dug in for the night along the east slope of the hogback about 3,000 yards from the beach which it had left almost five hours earlier.27  On the left flank, west of the hogback, BCT 17-3 had made about the same progress. Somewhat behind in the early stages of the advance, BCT 17-3 caught up with BCT 17-2 when the latter was stopped. It was now 2030. The two battalions, BCT 17-2 on the right side of the hogback and BCT 17-3 in the west arm of the valley, were abreast of each other, with the enemy in front of them firing from the heights that guarded the passes to Holtz Bay and Sarana Bay and from the ridges on both sides of the valley. At the request of BCT 17-3 the 105mm. guns back at the beachhead delivered a concentrated fire against the high ground at the head of the valley, and the battalion then attempted to resume its advance. But as soon as the artillery fire ended and the troops began to move forward, the Japanese again opened up. BCT 17-3 again halted and dug in somewhat ahead of BCT 17-2 on the other side of the hogback.28  

While the two battalions had been moving up Massacre Valley, two small detachments had been sent out on each flank to secure the ridges, Gilbert Ridge on the right and Henderson Ridge on the left. One of these detachments, a platoon of the 7th Reconnaissance Troop, landed on Alexai Point about four miles east of Massacre Bay Beach and halfway out the peninsula toward Chirikof Point. It was assigned the mission of establishing an outpost line across the peninsula from Alexai Point to Sarana Bay, of reconnoitering "to the west to include area between Lake Nicholas and Massacre Bay," and of afterwards reconnoitering the peninsula to the east, in the direction of Chirikof Point. It was to "make contact and co-ordinate efforts" with a platoon of the 17th Infantry in the pass between Sarana Bay and Massacre Bay. After landing on Alexai Point, the platoon from the 7th Reconnaissance Troop was out of contact with the main landing force for

[289]


two days. During this time it encountered none of the enemy and played no direct part in the battle. Had the Japanese attempted to infiltrate across Gilbert Ridge, the platoon might have played a more active role, even though its position was far to the east of any probable point of counterattack. The other platoon was from Company F, 17th Infantry. Reinforced with a light machine gun section and a 60-mm. mortar squad, this platoon had moved east along the shore of Massacre Bay and up into a steep pass leading over Gilbert Ridge to Sarana Beach. Its mission was to seize this pass and the "high ground along right flank" (i.e., Gilbert Ridge) to establish defensive positions in the Sarana end of the pass from which Sarana Beach and Lake Nicholas could be fired upon, and to "clear the ridge of enemy." It was to assist BCT 17-2, "if practicable," in the capture of the important pass at the head of Massacre Valley by firing on enemy installations at the western end of Gilbert Ridge. The platoon climbed all night and on the morning after D-day it was on the Sarana Beach side of the mountains. There it was discovered by the Japanese. For two days the men fought off strong enemy patrols, while they struggled westward along Gilbert Ridge. They finally managed to rejoin the main force in Massacre Valley near the point where BCT 17-2 had established itself on the night of D-day. The full story of their experience is one of the minor epics of Attu.29  Meanwhile, the detachments which had been sent out on the left flank of the Southern Landing Force to secure Henderson Ridge and the country beyond had likewise run into difficulty. A platoon of Company I, 17th Infantry, which on landing at Massacre Beach had been dispatched to secure the valley side of the ridge found rough going along the lower slopes. When fog and darkness finally halted the platoon, it had reached a point approximately 700 yards short of the position where BCT 17-3 had established itself in the valley. A week later, on 18 May, the platoon was only some 500 yards beyond its original positions. Further out on the left flank, behind Henderson Ridge, Company F of the 32d Infantry ran into several blind alleys after reaching its first objective, Temnac Cove. Although delayed by having landed farther to the east than it should have, Company F reached Temnac Cove by nightfall of D-day. There an enemy outpost was discovered and destroyed before the defenders were aware of the approaching Americans. Company F reported that its first mission, that of clearing the Temnac Cove area, was accom-

[290]


plished. The next morning the company proceeded on its second mission, to move northeast toward Holtz Bay, under orders to clear, as it went, all enemy installations from the flank of the main landing force advancing up Massacre Valley. It made no progress, however. Everywhere it turned it found itself in a cul-de-sac, and finally General Brown ordered the company to retrace its steps to Massacre Beach.30  

When General Brown went ashore toward the end of D-day the tactical situation was far from clear, but what information was available would not have indicated that a long drawn-out struggle was in prospect. The Southern Landing Force appeared to be close to its immediate objective-the passes leading from the head of Massacre Valley to Holtz Bay and Sarana Bay. BCT 17-3 reported that its position was about 600 yards short of the Holtz Bay pass, and BCT 17-2 was believed to be within 1,000 yards of the pass leading to Sarana Bay. There was a possibility that, on the northern front, BCT 17-1 had reached its first objective, Hill X. BCT 32-2, except for Company F, had not yet been committed, and the other two battalions of the 32d Infantry were due to arrive from Adak within twenty-four hours. If additional reinforcements were needed, General Buckner was willing to release for this purpose the 4th Infantry, which was being held in readiness to occupy Shemya Island as soon as Attu was secured.31  Everything considered, it would not have been unreasonable to suppose that within a few days the island would be taken.

Unfortunately, things were not entirely as they seemed. When the situation began to unfold on the morning of D plus 1, it became apparent that a long, hard fight was in store. In Massacre Valley, BCT 17-2 was at least 500 yards further from its objective than it had supposed, and BCT 17-3, apparently having mistaken a blind valley (Zwinge Valley) for the Holtz Bay pass, was a good 2,000 yards south of its immediate objective. Since neither Gilbert Ridge nor Henderson Ridge had been cleared, both battalions came under fire from each flank as well as from the front. BCT 17-2, which had been ordered to consolidate and hold its position with mission of blocking the Sarana-Massacre pass, thus found it necessary to move forward over very rough terrain in the face of heavy fire. Among the casualties was the regimental commander, Col. Edward P. Earle, killed by machine gun fire while with one of the forward elements. His death was a severe blow

[291]


to the 17th Infantry, and, in appointing Col. Wayne L. Zimmerman to take his place, General Brown deprived himself of the services of an extremely able chief of staff. At the end of D plus 1 the battalion was in position to block the pass, but the Japanese defenses were still intact. To the left of the hogback, BCT 17-3 managed to move forward to the rising ground at the mouth of the Holtz Bay-Massacre (Jarmin) pass, where it was pinned down. Frontal attacks against the mouth of the pass failed completely on each of the two days following, although BCT 17-3 now had the support of BCT 32-2 (minus Company F). By the middle of D plus 3 (14 May), it appeared that the Massacre Valley assault was stalemated.32  On the north side of the island, BCT 17-1 was faring no better in its attack against Holtz Bay. The height reached by the battalion on the night of D-day turned out to be some 900 yards short of Hill X, which the Japanese had occupied during the night. After bitter day-long fighting, BCT 17-1 won its objective during the early evening of D plus 1, but another two days passed before it could force the stubbornly resisting Japanese off the reverse slope and the shoulders of the hill. At the end of D plus 3, the battalion, now joined by BCT 32-3, was only 300 yards nearer Holtz Bay, while the Provisional Battalion, which had moved down into the canyon leading to the rear of the Holtz Bay positions, had been bottled up in the same position for nearly three days, about a mile from the mouth of the canyon.33  In a memorandum for Admiral Rockwell, General Brown summed up these first days of the battle in the following words:

Reconnaissance and experience of four days fighting indicates Japanese tactics comprise fighting with machine guns and snipers concealed in rain washes or in holes or trenches dug in each side and at varying heights of hill along narrow passes leading through mountain masses. These positions are difficult to locate and almost impossible to shoot out with artillery. They produce casualties in excess of casualties which can be returned. Number of machine gun positions out of proportion to estimated enemy strength. In addition, small infantry groups are dug in high up on sides of pass parallel to axis of approach through pass as well as all commanding terrain features in passes. Impossible to approach positions on sides of pass from above due to precipitous slopes more or less snow covered and extremely slippery footing. Progress through passes will, unless we are extremely lucky, be slow and costly, and will require troops in excess to those now available to my command. 34  

[292]


After repeated inquiries on the part of General Brown, the two battalions of the 32d Infantry (BCT 32-3 and BCT 32-1) that constituted the force reserve had arrived at Attu on 13 May (D plus 2). Although their arrival greatly eased the solution, General Brown and his staff were of the opinion that further reinforcements were necessary, specifically, part of the 501st AA Battalion, and a few miscellaneous units of the 7th Division, which were at Adak, and the 4th Infantry, which General Buckner had promised to make available for tactical employment on Attu if needed. At a conference on board the Pennsylvania, on 15 May, Admiral Rockwell was not at first convinced that these additional troops were necessary, but he finally agreed to forward General Brown's recommendations to Admiral Kinkaid with his concurrence. 35  Upon his return from the conference aboard Admiral Rockwell's flagship, General Brown drafted a message for General DeWitt, who was then at Adak, informing him that he had "made frequent attempts to . . . procure additional troops but without success," and requesting General DeWitt's assistance in the matter. 36  

Both Admiral Kinkaid and Admiral Rockwell were becoming increasingly concerned over the exposed position of the naval support forces. Japanese submarines were in the area. A torpedo had just missed the Pennsylvania on D plus 1 and on the morning of 15 May, soon after General Brown had returned on shore from his conference with Admiral Rockwell, four torpedoes narrowly missed one of the transports near the flagship. The other two battleships, Nevada and Idaho, had expended all their 14-inch high-capacity ammunition and, with their screen, had withdrawn to the northward to await orders. In view of the submarine threat, Admiral Kinkaid thought that the naval support group had tarried long enough in the dangerous waters of Attu. Accordingly, Admiral Rockwell had informed General Brown during the conference of 15 May that the ships would withdraw the next day, or in any event no later than the 17th. The continued requests for reinforcements, a long dispatch requesting large quantities of engineering and road building equipment, and the lack of any positive indications of a speedy breakthrough ashore persuaded Admiral Kinkaid that General Brown was bogged down. General DeWitt and General Buckner, whom Kinkaid consulted, agreed with him that it was necessary to relieve General Brown. Upon their recommendation, Admiral Kinkaid appointed General Landrum to take over the command of Attu. The new commander arrived on the scene during the afternoon

[293]


of 16 May and assumed command of the landing force at 1700, just as the fighting for Holtz Bay was reaching its final stage.37

General Brown's relief coincided with an advance of the Northern Force that broke the deadlock on Attu. Intense shelling by naval guns and bombardment from the air persuaded the enemy to start withdrawing from the West Arm area at Holtz Bay on 14 May, and on the next two days the Northern Force after making contact with the Provisional Battalion moved into the West Arm area and, against hot resistance, on to the high ridge that separated the two arms of the bay. The ridge itself was won during the night of 16 May. This placed the Northern Force (BCT 17-1 and BCT 32-3 ) directly in the rear of the Japanese defending the Massacre Valley pass, and on the morning of 17 May the Japanese began to withdraw toward Chichagof Harbor. The junction of the Northern and Southern Forces took place during the early morning hours of 18 May (D plus 7) , when a patrol from K Company of BCT 17-3 met the 7th Reconnaissance Company on the western slope of the Holtz Bay-Massacre Pass. 38

The Japanese withdrawal and the junction of forces marked the turning point of the campaign. Although nearly two weeks more of hard, costly fighting remained, the uncertainty and frustration of the first few days on Attu never recurred. It was a slow business taking the machine gun and mortar nests left on the heights by the retreating Japanese, but eventually the combined American force, reinforced with a battalion of the 4th Infantry, drew a net around Chichagof Harbor. 39  The end came on one frenetic night when most of the surviving Japanese, from about seven hundred to a thousand strong, charged madly through the American lines, screaming, killing, and being killed. The next day, 30 May, the enemy announced the loss of Attu, and units of the 32d, 17th, and 4th Infantry cleared out pockets of surviving

[294]


enemy troops as they advanced to occupy the Chichagof installations. Although mopping-up operations continued for several days, organized resistance ended with the wild charge of 29 May, and Attu was once more in American hands. 40

Out of a force that totaled more than 15,000 men before the campaign ended, 549 Americans had given their lives on Attu, 1,148 had been wounded, and about 2,100 had been taken out of action by disease and nonbattle injuries. Most of the nonbattle casualties were exposure cases, victims of the climate and weather and of inadequate clothing. Trench foot was the most common affliction. 41  The Japanese lost their entire force: approximately 2,350 enemy dead were counted and 29 taken prisoner. 42  The price of victory was high. In terms of numbers engaged, Attu ranks as one of the most costly assaults in the Pacific. In terms of Japanese destroyed, the cost of taking Attu was second only to Iwo Jima: for every hundred of the enemy on the island, about seventy-one Americans were killed or wounded.

Kiska-Grand Anticlimax

Before the guns had ceased firing on Attu, preparations got under way for the next moves against the Japanese in the Aleutians. An airfield was begun on Alexai Point, scene of one of the Attu landings, and on 30 May garrison troops and engineers landed on Shemya Island, thirty-five miles east of Attu, to begin construction of a bomber field there. Fighter strips were completed at both places before June ended, and in mid-July bombers from the new Alexai Point field made their first strike against Japan, a raid against the northern Kurils.

Even before the Attu landings took place preparations had been started for assaulting Kiska. 43  For this purpose on 4 May 1943 General DeWitt's headquarters had authorized the activation of an amphibious training force under General Corlett. Preliminary training was to be conducted at Fort Ord and San Diego by the Joint Staff that had planned the Attu operation, but advanced training at Adak was also to be provided. As a result of the Attu

[295]


experience and of revised estimates of the Japanese strength on Kiska, the assault force was doubled in size over that originally planned and among the additions were a mountain combat team, a regimental combat team from the Alaska Defense Command, and the hard-bitten First Special Service Force, all of them trained in the type of fighting that had developed on Attu. It was decided also, after the Attu campaign ended, to substitute the battle-tested 17th Regiment for one of the infantry units from California. By the end of July, about 34,000 Allied troops were assembled at Adak and Amchitka for final training in preparation for the assault on Kiska. Included among them was a Canadian brigade group numbering 4,800 officers and enlisted men, and about 700 men of the First Special Service Force were Canadians. 44  The enemy's strength on Kiska was estimated at from 9,000 to 10,000 men. Although some of the War Department planners favored postponing the operation, the joint Chiefs of Staff, upon the recommendation of the joint Staff Planners, gave their formal approval on 22 June. 45  General DeWitt and Admiral Nimitz designated 15 August as the target date. The force commanders, in conference at Adak on 30 July, decided that D-day ought to be postponed until 24 August to permit further training and regrouping of the battalion combat teams; but Admiral Nimitz was opposed to the delay, and D-day was definitely set for 15 August. 46

Unlike Attu, Kiska was subjected to a heavy preinvasion bombardment. Reinforced during June and operating from the new airfields, the Eleventh Air Force dropped a total of 424 tons of bombs on Kiska during the month of July. On 6 July and 22 July, strong naval task groups blasted the island with an additional 330 tons of explosives. 47  On 2 August a strong force consisting of two battleships, two heavy cruisers, three light cruisers, and nine destroyers carried out another bombardment, supported by seventeen bombers and eight fighters of the Eleventh Air Force. Over 200 tons of shells and bombs fell on Kiska on that day. Two days later, on 4 August, the Eleventh Air Force dropped a record-breaking 152 tons of bombs. Returning fliers claimed excellent results and reported "only meager and inaccurate" flak and small arms fire. Then, for several days while bad weather grounded

[296]


the Army bombers, destroyers of the naval blockading force continued the attack. On 10 August the Eleventh Air Force came back into the picture with another hard blow, and between then and D-day it dropped 335 tons of bombs on Kiska. 48  

Surprisingly enough, most pilots saw no signs of activity on the island; a few reported that they had encountered light antiaircraft fire. Earlier, there had been considerable success against Japanese submarines going to and from Kiska; then the enemy submarine traffic seemed to stop. The reports were the cause of considerable speculation at Admiral Kinkaid's headquarters. On one occasion, when someone raised the question whether the Japanese might not have been evacuating troops by submarine, Admiral Kinkaid, with a laugh, said he'd be glad to provide free transportation to Japan for half their garrison. His serious opinion was that the enemy had taken to the hills, as they had on Attu, and after wrecking all installations not already destroyed by the air and sea bombardment, were digging in for a last stand back from the beaches. The possibility of evacuation was not ignored, however. Shortly before D-day the suggestion was made that a small reconnaissance party be landed on Kiska by submarine in order to clear up the situation, but it was vetoed by Admiral Kinkaid. His position was that if the Japanese were still on the island the assault force was ready for them, but a reconnaissance party might be wiped out; that if the Japanese were not there, a landing would be a "super dress rehearsal, good for training purposes," and the only foreseeable loss would be merely the let-down experienced by the highly keyed troops. 49  With D-day only a few days away, Admiral Kinkaid decided to let the assault proceed as planned, without sending in a reconnaissance party.

Early on the morning of 15 August General Corlett's forces made a feint toward the south shore of Kiska and then landed on the north and west sides of the island. Not a shot was fired as the troops came ashore and moved up into the mist-shrouded interior. As on Attu, complete surprise seemed to have been achieved. All through the first night and the next day, and for several days afterward, American and Canadian patrols probed deeper into the island, occasionally hearing the noise of gunfire, but never encountering any Japanese. Kiska was an uninhabited island. The only guns that fired were those of friend against friend, and partly on that account casualties ashore during the first four days of the operation numbered 21 dead and

[297]


121 sick or wounded. The Navy lost 70 dead or missing and 47 wounded when destroyer Abner Read struck a mine on 18 August. 50

The entire enemy garrison had slipped away unseen, as the remnants of the Japanese Army on Guadalcanal had done six months earlier. To make the embarrassment complete, the Kiska evacuation had been carried out as early as 28 July, almost three weeks before the Allied landings. The original plan of the Japanese Imperial General Headquarters had been to withdraw the garrison gradually, by submarine, but this scheme had been given up late in June because of the loss or damaging of most of the submarines assigned to the operation and because of some anxiety that, by weakening the garrison over a prolonged period, the operation might fail. It was then decided to evacuate the entire force at one time, in one movement, using cruisers and destroyers as transports. The date, at first fixed for early July, was postponed until 28 July.51 Between then and D-day Kiska had been under attack and close surveillance by American naval units and the Eleventh Air Force, but the reports of flak and Japanese activity when there was none, which inexperienced observers brought back, obscured all the evidence from which the proper deduction might have been drawn. 52  Surprise was achieved, but it was not the Japanese who were surprised.

The retaking of Attu was the high point of the war, as far as it concerned Alaska. Kiska was anticlimactic, and what happened afterward was chiefly a matter of tying up the loose threads of unfinished business: of deciding upon the role that Alaska and the Aleutians could play in defeating Japan, and of making the organizational changes that the situation seemed to require.

In ridding the Aleutians of Japanese invaders, the objective had been partly to eliminate a potential military threat, but mostly to eradicate a psychological blot. As for using the western Aleutians as steppingstones to Japan, that idea had still to receive official imprimatur. General DeWitt and others had from time to time urged an assault by this route, but commitments to other theaters, and the desire of the Soviet Union not to have its neutrality

[298]


with Japan compromised, had precluded acceptance of the idea. 53  With the Aleutians cleared, and about 144,000 American and Canadian troops in the Alaska-Aleutians area, a reconsideration of the strategic role of that area seemed to be in order. 54  

An invasion of Shumushu and Paramushiro, the northernmost of the Kuril Islands, was the substance of a plan that General DeWitt submitted to General Marshall early in August. This plan contemplated using the combined forces that had been engaged in the Attu and Kiska operations, after they had been raised to a strength of approximately 54,000 men. It proposed a reinforcement of the Eleventh Air Force in heavy and medium bombers in order to provide the necessary long-range air support, and recommended organizing a North American theater to carry out the invasion in April or May 1944. Neither the Army's Operations Division, nor the Navy as represented by Admiral King and Admiral Nimitz, nor the joint Staff Planners saw any immediate possibility of implementing the plan. Pacific Fleet forces were fully committed to operations in the Central and South Southwest Pacific Areas; the creation of a North American theater in the North Pacific was not acceptable to Admiral King; and the seizure of the two northern Kurils without immediately pressing on toward Japan proper would, according to the Operations Division, place the forces in a position as hazardous as that of the Japanese in Kiska had been. 55  In deciding against an invasion of the northern Kurils in early 1944, the joint Chiefs nevertheless held the door open to the possibility of the situation in the North Pacific being altered, perhaps in favor of invading the Kurils, by the entry of the Soviet Union into the war against Japan. The War Department accordingly directed General Buckner to co-operate with Admiral Kinkaid in planning an assault on Paramushiro with the target date, for planning purposes only, set for the spring of 1945. 56  By the time the target date rolled around, the

[299]


American forces on Iwo Jima and Okinawa were only half as far away from Tokyo as Paramushiro was. 57  

After August 1943, whatever plans were discussed or even drawn up for assaulting the Kurils or Japan proper, Alaska like the Caribbean area and the Atlantic bases, was actually called upon to retrench, to reduce the strength of its garrison and curtail facilities. Within two weeks after the reoccupation of Kiska, four bomber squadrons of the Eleventh Air Force were designated for withdrawal, a reduction of the garrison strength to 80,000 by July 1944 was planned, the question of reducing the category of defense was brought up, and the reorganization of the Alaska Defense Command into a separate department was proposed. The cutback of bomber strength was carried out in September. In the following month, October, the separation of the Alaska Defense Command from the Western Defense Command and its redesignation as the Alaskan Department was announced, effective 1 November. 58  Both the proposed cut in garrison strength and the lowering of the category of defense were approved by the joint Chiefs before the end of October. 59  By the end of 1943 Army forces in Alaska had been reduced to about 113,000 men and General Buckner was notified to prepare for a further cut-to a total strength of 50,000. 60  This figure was approximately reached by the end of 1944. In spite of the fact that at this time the possibility of staging an offensive via the Aleutians began to revive, the process of retrenchment continued, and no serious consideration of reversing the trend was entertained. Any danger to Alaska and the Western Hemisphere had long since disappeared.

[300]


page created 30 May 2002


Endnotes

Previous Chapter        Next Chapter

Return to the Table of Contents