A related decision on 17 September committed the III Corps’ final reserve to the Ulithi landing. The task was assigned to the Western Attack Force, which was ordered to seize Ulithi with “available resources.” Over General Smith’s advice, Wilkinson chose to commit the entire 323d RCT, the 81st Division’s other maneuver element. The 321st subsequently and successfully occupied an undefended Ulithi while reserves were sorely needed at Peleliu.
By 20 September, the 81st Division had defeated or cornered all survivors of Angaur’s 1,400 defenders. The 81st’s commander declared Angaur secure. He tasked his 322d RCT to complete the mop-up, and reported to General Geiger that the 321st RCT was available for further operations. The lack of enough troops to begin encircling Umurbrogol was no longer an obstacle.
The other obstacle to reinforcing the division on Peleliu and encircling the Pocket lay in the thinking of General Rupertus, who clung to a belief that his Marines could do it without help from the Army. The III Corps plan tasked the 81st Division to reinforce the Marines in seizing Peleliu and then to relieve the 1st Marine Division for the mop up, but the general continued to exhort his commanders to “hurry up.”
Earlier, General Rupertus and Colonel Puller had shrugged off a suggestion from the 5th Marines’ “Bucky” Harris that they take a look at the Umurbrogol Pocket from the newly available light planes of Marine Observation Squadron 3. Harris’ own aerial reconnaissance, made immediately after those planes arrived on 19 September, had altered his view of the Umurbrogol from sober to grave. It convinced him that attacking the Pocket from the north would be less costly than the originally planned and ordered attempts from south to north. Both Puller and Rupertus responded to Harris that they had their maps.
[Illustration: _Once the troops entered the Umurbrogol Mountain, they found sinkholes and difficult terrain much as pictured here. Japanese soldiers in the caves and heights above could fire at will at the Marines, who were like so many “fish in a barrel.”_
Department of Defense Photo (USMC) 108432 ]
The prelanding scheme of maneuver was built on the tactical concept that, after capturing the airfield, the 1st Marine Division would push north along a line across the width of the main or western part of the island. Once abreast of the southern edge of Umurbrogol, that concept and maneuver scheme were reflected in a series of four west-to-east phase lines, indicating an expected linear advance, south to north. Clearly, it was expected that the advance along the flatter zones west and east of Umurbrogol would be at approximately the same pace as that along the high-central ground of Peleliu. Such thinking may have been consistent with Rupertus’ prediction of a three-day assault. Developments in Sabol’s sector to the west, and in the 5th Marines’ sector to the east, apparently did not change division-level thinking. Until additional forces became available, such a linear advance may have seemed all that was possible.