CHAPTER V

Study on Wound Ballistics-Bougainville Campaign

Ashley W. Oughterson, M.D., Harry C. Hull, M.D., Francis A. Sutherland, M.D., and Daniel J. Greiner, M.D.

The purpose of the wound ballistics study¹ conducted on Bougainville was to obtain information on the relative effectiveness of different weapons as casualty-producing agents. To obtain this information, a study was made of all battle casualties (living and dead) which had occurred in the U.S. Army Ground Forces on Bougainville Island from 15 February to 21 April 1944.

Though it was possible to obtain information on all casualties, living or dead, for the entire period from 15 February to 21 April, post mortem examinations were limited to the interval from 22 March to 21 April. The number of autopsies was further curtailed because the bodies of some of those killed in action were not obtained before decomposition was far advanced. It was also hoped to study the effect of U.S. Army weapons on the enemy dead. The character of the fighting resulting in multiple wounds by rifle, machinegun, grenade, mortar, and artillery fire made it almost impossible, however, to determine what weapon was responsible for death. Furthermore, because of delay in obtaining the Japanese dead, the state of deterioration frequently precluded post mortem examination. Also, during this period, it often required all the available personnel to perform post mortem examinations on U.S. Army killed-in-action casualties.

Since the effect of weapons may be observed on the living as well as the dead, a clinical appraisal especially with regard to end results was needed. Furthermore, the relative effect of weapons may be greatly influenced by the quality of medical care. For this reason, the ballistics team after completing the study in the forward area followed the patients through the hospitals of the rear echelon.

The battle casualties studied may be divided into two large groups: The killed in action and the wounded in action.

¹In accordance with instructions from The Surgeon General, 21 January 1944, a team was organized for the purpose of conducting a study on wound ballistics. This team included Col. Ashley W. Oughterson, MC, Surgeon; Lt. Col. Harry C. Hull, MC, Surgeon; Maj. Francis A. Sutherland, MC, Surgeon; Maj. Daniel J. Greiner, MC, Pathologist; Sgt. Reed N. Fitch, T4g. Charles J. Berzenyi, and T5g. Charles R. Restife. The team was organized to participate in the contemplated New Ireland operation and was ordered to Guadalcanal for training and organization. The New Ireland operation was cancelled, and the team was then ordered on detached service with the XIV Corps on Bougainville and reported there on 22 March 1944.

282

Killed in action.-Those killed in action prior to 23 March were recorded in the graves registration files. While some of these records were excellent, many were inadequate. Information on the circumstances attending death, such as type of missile, distance from burst, terrain, time, and type of protection, was supplemented by personal interviews with the medical officers and aidmen or with comrades who, during the action, had seen the soldier killed or had seen him before he expired. This information is better obtained by personal interview than by questionnaire because the circumstances attending death are so varied. In order to obtain reasonably accurate data, evaluation of the situation by trained and interested personnel is necessary at the time of interview. Subsequent to 23 March 1944, all the dead were brought to the 21st Evacuation Hospital which was located near the cemetery. Here, excellent facilities and assistance for post mortem examinations were available. This work was carried on by the pathologist who was assisted by a clerk and a photographer. When the number of autopsies exceeded 10 or 12 per day, additional assistance was provided by the surgeons. A few additional post mortem examinations were obtained on those wounded in action who died later in hospitals of the rear echelon. A card index was kept on all wounded, and this was checked for death against the records of the hospitals in the rear echelon.²This check was made at a later date, and for the majority of patients, a period of 1 to 4 months had elapsed since they were wounded; hence, there is reason to believe that all or nearly all of the dead are recorded in this study.

Wounded in action.-The wounded in action fell into three groups: (1) The more seriously wounded who were evacuated from Bougainville, (2) the relatively minor wounds treated in the clearing stations or hospitals and returned to duty in 1 to 3 weeks, and (3) the very minor wounds and abrasions returned to immediate duty from the battalion aid and collecting stations. This latter group was not studied. The second group, of minor wounds treated and returned to duty from the clearing stations and hospitals, were studied in detail, as were those evacuated from the island. Factors relating to ballistics in the wounded in action were obtained by questionnaire and by personal interview. The personal interview was undoubtedly superior, but since these troops were still in battle it was sometimes impossible to obtain an interview with an eyewitness. When emergency medical tags and hospital records were checked with eyewitness accounts, many discrepancies were found as to the weapon, the distance, what the soldier was doing, and the exact circumstances surrounding his injury. Allied officers and enlisted men were questioned regarding effectiveness of enemy weapons and tactics, as well as their own. Questions were also asked regarding the construction of pillboxes and the use of camouflage with reference to their effectiveness as a means of protection.

²Throughout this chapter, hospitals in the rear echelon refer to those on Guadalcanal, Espíritu Santo, and New Caledonia.-J. C. B.+

283

FACTORS PECULIAR TO THE BOUGAINVILLE CAMPAIGN

Geography

Bougainville Island is in the northernmost part of the Solomon Islands group, lying between latitudes 50°28' S. and 5°51' S. It is approximately 130 miles long with an average width of 30 miles. It is a tropical island of volcanic origin with a backbone of rugged mountain ranges. Behind the Empress Augusta Bay sector, the Crown Prince Range rises to a height of 6,560 feet with an active volcano, Mount Bagana. The Empress Augusta Bay and Torokina Point sectors present a low sandy shoreline with heavy surf. The south shore of this island has very little coral, and behind the shoreline a sandy alluvial plain rises gently to the foothills of the Crown Prince Range, about 4,000 yards inland. Near the shore are some lagoons and in the region of the Torokina River extensive swamps. The subsoil of the plain is black volcanic sand providing good drainage. The rainfall which is fairly uniform throughout the year averages approximately 11 inches per month. The typical heavy tropical showers wash and erode the hillsides and make constant road maintenance a necessity.

The Empress Augusta Bay beachhead was virgin jungle except for a small coconut plantation on Torokina Point. The elaborate system of roads shown in the situation map (fig. 159) had all been built since the initial landing during the first week of November 1943. At the time of the enemy attack on 8 March 1944, this system of roads was nearly completed except for a section of the perimeter road connecting the Americal and 37th Division sectors. The perimeter at its greatest depth was carried along the high ridges of the foothills, and this extremely rugged terrain presented a major problem in evacuation where roads were not present or were under fire. This road system alone played an important role in saving the lives of many casualties which might otherwise have been lost. However, the problem of evacuation of wounded within the perimeter was simple when compared to the difficulties encountered in evacuating men wounded on patrol. Patrols constantly covered this rugged terrain beyond the perimeter for distances of 1,000 to 8,000 yards. Even a 1,000-yard carry over these ridges and draws was exhausting to both the litter bearers and the patient.

Medical Installations and Routes of Evacuation

The medical installations and routes of evacuation (fig. 153) were better developed on Bougainville at the time of the attack than for any other island campaign in the South Pacific. This was due to the fact that the beachhead had been developed steadily over a period of 4 months before the Battle of the Perimeter began.

284

FIGURE 153.-One of the routes of evacuation between the clearing station of the Americal Division and the 21st Evacuation Hospital.

Two-way all-weather roads made all parts of the perimeter easily accessible with one exception. This one sector lay near the boundary line between the Americal and 37th Divisions, where the perimeter road had not been completed (fig. 154). Furthermore, the roads were kept open throughout the battle except on Hill 700. The one-way all-weather road over very rugged terrain leading to the latter Hill was for a time under enemy fire, and as a result a difficult litter carry of 1,200 yards was necessary during the attack. Later at this point, and at others where sporadic fire was encountered, half-tracks were used for evacuation, and patients were then transferred to jeep ambulances and taken to the hospital. The greatest distance from the front-line to a clearing station was found on the Americal sector at the mouth of the Torokina River which was approximately 10, 000 yards over a good road. Figure 155 is an illustration of the type of road which existed outside of the perimeter area.

Owing to this excellent system of good roads, the majority of patients arrived at the hospitals within 3 hours, and frequently within an hour. A sample of 142 patients showed that 87 percent were on the operating table within 3 hours. Patrol missions presented the most difficult problems of evacuation. Small patrols, frequently no larger than a platoon, were so numerous that it was impractical to send a medical officer with each one. Larger combat patrols were usually accompanied by a medical officer. On only one occasion, however, was a patrol large enough to warrant the use of a portable surgical hospital. As a consequence, some patients who were wounded on patrol did not reach the hospital until after 24 to 48 hours had elapsed. However, every effort was

285

FIGURE 154.-Perimeter road near junction of Americal and 37th Divisions. A good route of evacuation over difficult terrain built by the 117th Engineer Combat Battalion.

FIGURE 155.-Roadway along the Laruma River, outside the perimeter.

286

made to reduce delay to the minimum and to provide surgery at the earliest possible moment.

The medical installations available for the Bougainville campaign were more than adequate. The clearing stations of both the Americal and 37th Divisions had been augmented with additional surgical equipment before the hospitals were established on the beachhead. The 31st Portable Surgical Hospital had been assigned to the Americal Division and the 33d Portable Surgical Hospital, to the 37th Division. Owing to the fact that more adequate medical facilities became available later, the portable surgical hospitals were not necessary, although they were both utilized. The 52d Field Hospital was utilized for the care of service troops and functioned chiefly as a station hospital for the island. The 21st Evacuation Hospital (figs. 156 and 157), an affiliated unit from the University of Oklahoma, Norman, Okla., had an exceptionally well qualified staff, including the various specialists. The construction of this hospital was completed on 8 March 1944; however, the hospital had functioned for a limited number of patients since 15 February. The normal capacity of the 21st Evacuation Hospital was 750 beds with facilities available for an additional 250 beds (fig. 158). Casualties from all combat troops were cared for at this hospital. Since the 21st Evacuation Hospital was situated only 4,000 yards from the frontlines at the nearest point of attack (forward of some artillery batteries), the majority of the seriously wounded patients were sent directly to the hospital to avoid delay at the clearing stations.

FIGURE 156.-Ward area of the 21st Evacuation Hospital on Bougainville.

287

FIGURE 157.-Underground operating room of the 21st Evacuation Hospital on Bougainville. There was a similar operating room above-ground providing space for eight tables.

FIGURE 158.-Interior of underground ward, 21st Evacuation Hospital on Bougainville. Space was provided for 120 litter patients. This would have been inadequate if shelling had been heavy.

288

All patients from the island were evacuated through the 21st Evacuation Hospital. Nearly all patients evacuated to the rear were sent by air transport to Guadalcanal and were cared for there in one of three 500-bed station hospitals. Patients requiring a long period of convalescence were evacuated from Bougainville by ship or air transport to Espíritu Santo and to New Caledonia.

Allied and Japanese Forces

Allied forces on Bougainville were concentrated in the Empress Augusta Bay beachhead. The perimeter line of defense had been extended previously in three phases until, by the time of the Japanese attack on 8 March 1944, it enclosed about 20 square miles and was approximately 22,000 yards in length. The total strength within this perimeter as of 31 March 1944 was 60,583. Included were 11,220 Navy and Marine personnel and civilians. The few casualties from these groups were due mostly to shelling and bombing and are not included in this study. The casualties included in this study were derived, therefore, from a total strength of 49,363. Of this number, 40,404 were U.S. Army Ground Force combat troops of which 27,831 constituted the 37th and Americal Divisions. The remainder of the ground force combat troops were attached to the XIV Corps and the 25th Regimental Combat Team. Allied forces other than U.S. troops, chiefly Royal New Zealand Air Force and Fijian Infantry, numbered 3,424. It should be noted that, of these forces, the number actually involved in combat was comparatively few. This number could not be ascertained except for certain specific engagements. The perimeter line of defense was divided between the 37th and Americal Divisions although other forces were used in the line at various times. The Fijian troops, among whom there were a considerable number of casualties, were used chiefly on patrol missions.

Immediately before the attack, the effective strength of the Japanese Army and Navy forces on Bougainville numbered about 27,000. Of these, about 18,000 were believed to be Army combat troops. The remaining strength consisted of Army and Navy antiaircraft, base, service, and labor troops. No surface ships had been observed in the Bougainville area since mid-November 1943 and whatever supplies were brought in were carried by submarine or barge. With the exception of small arms ammunition, there was evidence that the enemy was short of basic supplies. Although elements of the 17th Division (one battalion each from the 81st and 53d Infantry Regiments) were identified in the Torokina area, the brunt of the attack was borne by the Japanese 6th Division. The backbone of the enemy's strength was the 13th, 23d, and 45th Infantry Regiments (fig. 159). These units were supported by the 6th Field Artillery Regiment (2d Battalion) elements of the 4th Heavy (Medium) Artillery Regiment, as well as miscellaneous mortar, artillery, engineer, and road construction units. The 1st Battalion, 13th Infantry (minus one company), was to be the division reserve. The total strength of these units actually in combat in the Battle of the Perimeter was believed to be only slightly more than 10,000.

289

FIGURE 159.-Situation map. Disposition of principal enemy units, 29 February 1944.

The enemy forces faced great difficulties of transportation in the maneuvering of various units, especially heavy artillery, into positions favorable for attack. This had to be accomplished over the most rugged type of terrain at great expenditure of manpower. Finally, they attacked with almost no air support.

Description of the Weapons Commonly Employed by the Japanese³

In the period under study, Japanese weapons accounted for 1,569 casualties, including killed and wounded. Table 57 is a breakdown of the type of Japanese weapons responsible for 1,569 Allied casualties.

Estimates based upon captured weapons indicate that the ratio of 6.5 mm. (caliber .256) to the 7.7 mm. (caliber .303) rifle was approximately 4 to 1.

³A complete description of Japanese ordnance is contained in chapter I, pp. 4-35.

290

TABLE 57.-Japanese weapons responsible for 1,569 Allied casualties

Furthermore, of the smaller caliber (6.5 mm.) weapon, roughly 90 percent were "long," 7 percent "short," and 3 percent "medium" types.

The almost complete absence of muzzle flash in the Model 38 (1905) is a characteristic commented upon favorably by U.S. soldiers. Since the latest Japanese rifle, Model 99, did not possess this feature, it was apparently considered unimportant by the enemy.

Most commonly employed by the enemy at Bougainville, in a ratio of approximately 4 to 1, were the Model 96 (1936) 6.5 mm. light and the Model 92 (1932) 7.7 mm. heavy machineguns. Extremely rare was Model 11 (1922) 6.5 mm. light machinegun ("Nambu Keiki") among the 200 captured machineguns. Closely resembling the British Bren light caliber .303 model, the Model 96 (1936) 6.5 mm. light machinegun was considered an excellent weapon by American officers.

Wounds ascribed to the mortar at Bougainville in many instances were actually produced by the grenade discharger. Mistakenly called the knee mortar, this weapon, because of its accuracy and efficiency, had earned the respect of the American combat troops and was more feared than any other Japanese weapon. If the "knee mortar" was grouped with the other types of captured mortars, it was found to constitute approximately 90 percent of the total. Among the conventional mortar types, the ratio of the 81 mm. to the 90 mm. was about 3 to 2. A total of 96 mortars were captured, only one of which was the 90 mm. Model 97 (1937).

Because it could he thrown by hand, fired from a grenade discharger, or used as a rifle grenade, Model 91 (1931) hand grenade, "Kyuichi Shiki Shuryudau," was a useful, versatile, and frequently employed weapon. Model 97 (1937) hand grenade was similar to Model 91 except that it had no propelling charge and could not be fired from a grenade discharger. It was carried by all Japanese frontline troops but was said to have poor fragmentation, the fragments being small and of short range. The effective range from the burst was estimated at 5 yards and the danger zone, 30 yards.

In the plan to neutralize and seize the three Torokina airfields, the artil-

291

lery support was the most extensive yet employed by the enemy in the South Pacific. The Japanese were able to transport a considerable number of heavy weapons through dense jungle and over exceedingly rough terrain to positions overlooking the U.S. perimeter. Assuming all units at full strength, an order of battle indicates that the maximum number of weapons available to them was 136. Actual observation suggested the presence of approximately 40 or 50 pieces.

With the exception of the 10 and 15 cm. pieces, all weapons were of pack type and were undoubtedly carried by hand. Possibly the 150 mm. howitzers may have been dismantled also, as some of these were reported on Mount Bagana. These weapons were brought by water to Koaris and thence by road to the vicinity of Hills 500 and 501. Limited use of horses was reported on the Kahili-Empress Augusta Bay track. Apparently there was no serious shortage of ammunition by Japanese standards, fire having continued intermittently from some positions for 3 weeks. Considerable quantities of ammunition were generally found with the captured weapons.

Principal targets were the airstrips, supply and command post areas, road junctions, and the tank areas. Massing of fire was not utilized and gunfire seemed independent. The heaviest concentration occurred in the early morning and evening hours. On 23 March, in less than 2 hours, 70 rounds fell on the Piva airfields. After the first 2 days of attack, during which some parked planes were destroyed, rarely in a single day did more than five or six shells fall on these same airfields. Difficulties inherent in jungle warfare precluded the use of artillery in close support of attacking Japanese infantry. For this purpose, the Japanese relied principally upon 90 mm. mortar fire.

The Japanese employed at least thirty-five 75 mm. guns, Model 41(1908) and Model 94 (1934), the former predominating. These pieces were situated north and northeast of the perimeter. Four 150 mm. howitzers were located on the northeast and east and two 105 mm. howitzers on the east near Hill 501. Mortar fire received was principally from the north and northwest sections. The greatest concentration of fire in any one day was 200 rounds. In contrast to the experience during the weeks after the landing in November, the proportion of "duds" was remarkably low. Observers were able to identify by type of burst or by duds about 1,300 rounds received. Of these, 885 were 75 mm. shells and 130 were 150 mm. shells. Many types of artillery weapons were captured. The five most commonly encountered models will be described briefly.

Model 94 (1934) 37 mm. gun was designated "Kyuyon Shiki Sanjunana Miri Ho," and commonly called Sanjunana Miri Ho. It could be used both as an AT (antitank) and antipersonnel weapon, employing AP (armor-piercing), HE (high explosive), and shrapnel ammunition. This gun had a long, slender barrel measuring 66.5 inches in length. The effective range was 2,500 yards and the maximum range 5,000 yards. The total weight of the weapon in action was 714 pounds. The effective burst of the HE shell was said to be 10 yards with a zone of danger extending about 75 yards. Fragmentation tests

292

revealed that the 560 grams of metal in the shell broke into 490 fragments. Only 143 of these fragments were classified as lethal (average weight of lethal fragment being 3.1 grams).

The Model 92 (1932) 70 mm. howitzer (Battalion Gun), "Kyuni Shiki Hoheiho" was a horse-drawn infantry support howitzer. It weighed 468 pounds and could be handled by a 10-man section. It had an effective range of 1,500 yards and a maximum range of 3,000 yards. The estimated effective range of burst was 20 yards, and the area of danger was 200 to 300 yards.

Issued for use as an infantry regimental gun, the Model 41 (1908) 75 mm. mountain (infantry) gun was originally used as a field artillery pack gun. The effective range of this weapon was 2,100 yards, and it fired both HE and AP shells. With the long, pointed shell, its maximum range was 7,675 yards and with the ordinary shell, 6,575 yards. The total weight was 1,200 pounds. Its muzzle velocity was listed as 1,200 f.p.s. (feet per second). The shell had a probable effective burst of 20 yards with a danger zone of 300 feet.

The Model 96 (1936) 150 mm. mobile field howitzer has a range of 13,200 yards. The effective range of the shellburst was said to be 50 yards with an area of danger of 500 yards. The effect produced was that of blast and fragmentation.

The Model 98 (1938) 20 mm. AA/AT (antiaircraft, antitank) machine cannon was an all-purpose weapon. It was gas operated and semiautomatic or full automatic. The ammunition for this weapon was HE, tracer, and AP and was fed by a 20-round box magazine. This weapon was very maneuverable, weighing without wheels 836 pounds. The rate of fire was 120 rounds per minute. The muzzle velocity was 2,720 f.p.s. and the maximum ranges, horizontal 5,450 and vertical 12,000 feet.

A list of Japanese rifles, machineguns, mortars, grenades, and artillery weapons captured on Bougainville follows.

¹Models most frequently employed.

293

BOUGAINVILLE CAMPAIGN DURING SURVEY PERIOD (15 FEB.-21 APR. 1944)

The Allied beachhead was established during the first week of November 1943. The period before the Battle of the Perimeter was characterized by consolidation of the defenses of the airfields which were being used for attacking enemy installations in the Bismarck Archipelago and on Bougainville. By 15 February, the airstrips were completed and the perimeter established with the 37th Division on the left flank and the Americal Division on the right flank. From 15 February to 8 March, the perimeter defense was strengthened, and an extensive system of roads was further developed within the perimeter. During this period, patrols made contact with enemy forces moving into position north and east of the perimeter. Some artillery installations were discovered, and strong enemy positions were noted on Hills 1000, 1111, and 600 east of the Torokina River mouth (fig. 159). However, during this period, contact with the enemy was limited to patrol skirmishes and an occasional bombing raid at night.

The Battle of the Perimeter extended from 8 March to 24 March. The Japanese had laid plans for this offensive sometime around the turn of the year. Allied intelligence obtained information that the enemy attack was to be launched on 8 or 9 March, thereby permitting ample preparation for defense of the perimeter.

Enemy plan.-The three infantry regiments were to leave their respective lines of departure following an artillery barrage. This barrage was to commence at 0430 Y-day from the main strength of the 6th Artillery Regiment (fig. 159) located near Blue Ridge (mountain guns) and the medium field artillery (10 and 15 cm. field pieces) deployed near Hill 500. It appears that the 45th Infantry was to constitute the main thrust and was to strike Allied lines near the point where the Piva-Numa-Numa Road enters the perimeter (129th Infantry sector). Simultaneously, the 23d Infantry was to launch its attack from approximately 1,000 yards northeast of Hill 700 with the 3d Battalion on the left and the 2d Battalion on the right and the 1st Battalion in reserve. By the end of Y-day, the 3d Battalion was to have captured Hill 700 and the 2d Battalion was to have occupied Cannon Hill. These heights overlooked the Piva airstrip, and the main strength of the 23d Infantry was to have attacked the strip from the east while the 45th attacked from the west. The 13th Infantry was to attack Hill 260 and then join with elements of the 23d Infantry to proceed in the general direction of the airstrip.

The enemy's Torokina operation began on 8 March with preliminary artillery fire directed mostly on the Piva airstrips. Blue Force counterfire against hostile positions located in the general areas of Hills 1111 and 501 began immediately. The main Japanese drives began under the cover of darkness during the night of 8 March and the morning of 9 March at the three points on the perimeter. In the east sector patrol, contacts and fire fights took place in the vicinity of Hill 260. To the north on Hill 700, the Japanese

294

infiltrated through Allied lines and occupied the northwest slope of the hill. Blue Force counterattack reduced the Japanese positions, and the perimeter was reestablished. In the northwest sector, several fire fights occurred. The Japanese had occupied strong points on Hill 260 and severe fighting resulted in retaking these points, but by 11 March two Blue Force companies occupied Hill 260 with the exception of strong points on the southeast slope. Another attack on the northern sector was repulsed. Meanwhile, preparations for an enemy drive from the northwest continued. On 12 March, three major attacks from the northwest near the Numa-Numa Trail placed the Japanese within the U.S. perimeter. American tank-infantry teams reestablished the lines next day. The same Blue Forces on Hill 700 received and repelled the third attack on that position. On 15 March, another attempt was made by the Japanese to break through the sector held by the 129th Infantry. Tank-infantry counterattack again restored the perimeter. The next strike by the enemy was again from the northwest near the Piva-Numa-Numa Trail on 17 March. Although a 75-yard penetration was made for the third time, tanks and infantry drove the enemy back. For a week, the Japanese remained relatively quiet, regrouping their forces opposite the northwest sector of the perimeter. Smaller holding forces which were dug in were contacted on the other sectors. On 24 March, after a feeble attempt at laying an artillery barrage, the Japanese struck toward the Piva airstrips once more, penetrated the 129th Infantry lines, and again were driven back, losing 300 men and a field gun. On each occasion when penetration was made, the enemy succeeded in occupying pillboxes within the U.S. perimeter only to be dislodged with heavy losses.

The Japanese did not again attack in force after the repulse on 24 March and began a general withdrawal. Hill 260, however, was not evacuated by the enemy until 28 March. From 28 March to 22 April when this study was completed, contact with the enemy was limited to a few fire fights, patrol skirmishes, and occasional shelling of the airstrips. There were 5,522 Japanese dead counted between 8 March and 22 April. This, however, did not include all areas subjected to U.S. artillery fire.

Battle of the Perimeter

Operations on Hill 260⁴

The original garrison on Hill 260, a reinforced platoon from Company C, was attacked by a Japanese force of undetermined size at dawn on 10 March. The enemy generally occupied the area south of the outpost tree (fig. 160), and, from this date until the termination of the battle, the Japanese tried to increase their garrison and improve their positions on that side of the hill in order to secure observation for an all-out attack on the main line of resistance.

⁴Report, Lt. Col. Wm. J. Mahoney, Executive Officer, Headquarters, 182d Infantry, Americal Division.

295

FIGURE 160.-Focal point of entire Hill 260 battle. Banyan tree used as an Americal Division artillery spotting post. In the 20-day fight for the hill, 541 Japanese were killed.

The terrain was that of an elongated hill with moderately steep sides covered by rain jungle. The outpost tree (fig. 160), around which the heaviest fighting occurred, was one of a common variety of trees on Bougainville, the roots of which plus excavation make a very strong defensive position (fig. 161).

After the initial attack, the Japanese held the south end of Hill 260. They greatly increased the force which had made the original attack because they beat back the Allied attempt to storm the northwest, southwest, and southeast ridges of the hill during the period 11-17 March. Apparently, their main route of supply and evacuation was down the steep east side of the hill, then north clinging to the east side of the west bank bluff overlooking the Torokina River. This route was well concealed and in defilade and difficult to reach by fire. After the initial engagement, reinforcements were sent to secure the north side of Hill 260. The establishing of a perimeter there and the continual pressure on the Japanese positions completely neutralized the effect of the offensive action taken by the Japanese. The possession of Hill

296

FIGURE 161.-Banyan trees are common in the jungle on Bougainville and offer excellent protection. The outpost tree on Hill 260 was of this variety.

260 by the enemy would have jeopardized a considerable portion of the Allied main line of resistance.

From the outset, the problem on Hill 260 was one of ejecting the Japanese from the south end of the hill. Their positions were well dug in (fig. 162), and the various American assaults to take the hill were turned back with heavy casualties. Artillery and mortars were useful in blasting Japanese positions in the general area, but because of the proximity of American troops, prepared fires could not be used on the Japanese positions just outside the U.S. perimeter. Artillery was effective on the exposed southwest slope, and after a week's fighting the Japanese were pretty well removed from that area (fig. 163). But those in defilade on the southeast slope dug in and countered every American move. Various means were used to force the Japanese from their dugouts during the closing 10 days of the battle. It was obvious at that time that the Japanese garrison was considerably reduced in numbers although there was no corresponding lessening of firepower. Flamethrowers and gasoline ignited with thermite grenades reduced a few pillboxes (fig. 164). As late as the morning of 28 March, Japanese were seen near pillboxes on the southeast slope. On the morning of 28 March, three patrols were sent around the base of the hill to fire on the Japanese. When there was no fire, the Allied patrols investigated and found that the Japanese had evacuated. At 1246, 28 March, Hill 260 was secured. On the morning of 30 March, the 2d Battalion, 182d Infantry, was replaced on the hill by 1st Battalion, 24th Infantry.

297

FIGURE 162.-Jungle growth on Hill 260, showing protection afforded by trees.

FIGURE 163.-Hill 260 being shelled by Americal Division artillery fire, on 19 March. The firing continued for several hours at the end of which time it was believed that all enemy resistance had been neutralized. Note partial destruction of jungle growth.

298

FIGURE 164.-Enemy pillbox on Hill 260. The dense jungle growth has been entirely cleared away by artillery fire.

American forces engaged.-Companies B, E, F, G, and H plus one platoon of Company K, 182d Infantry, and Company G, 164th Infantry, actively took part in the action on the hill. All other companies in the regiment were in general support plus A and B Companies, 57th Engineer Combat Battalion; 246th and 247th Field Artillery Battalions; 82d Chemical Battalion-total, 1,350 men.

Japanese forces engaged.-Elements of the 13th and 23rd Infantry Regiments, both part of the 6th Division, were identified as taking part in the battle for Hill 260. It was estimated that 1,400 Japanese were involved in this action.

Table 58 lists the casualties sustained by the 1,350 U.S. troops engaged on Hill 260.

In comparison to the other two main thrusts by the enemy on the perimeter, there was more offensive action by U.S. troops on Hill 260. The enemy in the initial attack had captured and had managed to defend the outpost tree which was the focal point on the hill. Furthermore, the character of the terrain lent itself readily to defense and prevented the effective use of tanks.

The heaviest casualties were in the 182d Infantry with 800 troops involved (table 59).

Estimates of Japanese killed and wounded were difficult to make because of their practice of carrying away and burying their own dead. A total of 212 Japanese bodies were found by U.S. troops on Hill 260, and the Americal Division G-2 (intelligence) listed 541 Japanese as the total killed. The ratio of Japanese to U.S. troops killed was 7.6 to 1. In addition, many wounded

299

were seen going to the rear, and it is believed an entire battalion plus a number of supporting troops were virtually wiped out. The heaviest fighting occurred during the period 10-14 March and, as indicated later by prisoner-of-war reports, this engagement broke up the initial attack of the entire Japanese 13th Infantry Regiment on the Bougainville perimeter.

TABLE 58.-Distribution of 713 casualties among 1,350 U.S. Army troops engaged on Hill 260, by category

TABLE 59.-Distribution of 426 casualties among 800 men of the 182d Infantry engaged on Hill 260, by category

Operations on Hill 700⁵

The terrain here was mountainous but mostly second growth (fig. 165) rather than virgin jungle. The sides of the ridges were very steep (fig. 166) and at one point of the assault were almost precipitous. The action by U.S. troops was largely defensive but did involve the recapture of certain positions into which the Japanese had infiltrated. A one-way road leading along behind Allied lines for a time was under enemy fire (fig. 167), necessitating a difficult carry (fig. 168) for the litter bearers of more than a thousand yards.

⁵Report, After Action Operations, 37th Infantry Division, pt. II, G-3 Operations Narrative, 8 Nov. 1943-30 Apr. 1944.

300

FIGURE 165.-Partially cleared jungle growth on Hill 700. Through this draw, the Japanese made their approach to the hill.

FIGURE 166.-Precipitous hillside off the perimeter road. Grenades were rolled down this bank causing many casualties.

301

FIGURE 167.-Wounded being transferred from halftrack to jeep. Halftrack was used because road was under fire.

FIGURE 168.-Wounded soldier being helped down the side of Hill 700 by two medical aidmen.

302

On the morning of 8 March, the Japanese attack began with some artillery and spasmodic small arms fire which continued throughout the day. During the night of 9 March, boobytraps warned of attack followed by hostile fire from mortars and rifles. At dawn, it was found that at least one company of Japanese had occupied the north slope and crest of Hill 700 and had penetrated the Allied line to a depth of 75 yards over a 100-yard front. During the day, a counterattack by the 1st and 2d Battalions of the 145th Infantry regained several pillboxes on the south slope of the hill. One tank was used with fair success along the road which was under fire. On 10 March, the enemy retained possession of the crest of the hill in spite of continued ground action. Efforts to reach the Japanese positions on Hill 700 by engineer "polecharges," bangalore torpedoes, and bazookas were without avail and resulted in numerous casualties due to the excellent Japanese field of fire.

At 1700 hours on 10 March, a determined attack was made by U.S. forces who, in spite of intense enemy light and heavy mortar and artillery fire (fig. 169), succeeded in driving the enemy from the crest of Hill 700. Japanese concentrations coming up to reinforce this area were subjected to heavy bombing and artillery fire which was very effective. On 11 March at daylight, the enemy made a general assault on Cannon Hill held by the 3d Battalion of

FIGURE 169.-Japanese 75 mm. gun emplacement on Blue Ridge that was used by the enemy in their attack on Hill 700. Interior view; note the large window.

303

the 145th Infantry. The attack was repulsed with the exception of one pillbox gained by the enemy on Hill 700 (fig. 170). Japanese losses were reported as enormous with the enemy assault wave attacking over piles of their own dead. On 12 March after severe fighting, U.S. forces succeeded in driving the enemy from Hill 700. A total of 399 Japanese dead were counted within the wire on the crest and on the forward slope of the hill. On the night of 13 March, the enemy again attacked in the draw west of Hill 700. Searchlights were used successfully to reflect light from the overhanging clouds, and the attack was repulsed. After this date, only intermittent contact was made with the enemy in this area.

FIGURE 170.-Enemy dead killed while defending their position in a pillbox on Hill 700.

During the engagement on Hill 700, there were approximately 2,600 U.S. troops involved. Table 60 summarizes the various types of casualties among the 519 total casualties.

A total of 2,219 (719 counted, 1,500 estimated) Japanese were killed in action during the engagement of Hill 700. For this encounter, the ratio of U.S. dead (KIA plus DOW) to Japanese dead was 1 to 36.

The large number of enemy dead estimated rather than counted was due to the enemy custom of burying several bodies in one grave and also to the large number killed by U.S. bombing and artillery fire behind the lines, making it impossible to obtain an immediate count.

304

TABLE 60.-Distribution of 519 casualties among 2,600 U.S. troops engaged on Hill 700, by category

Operations on 129th Infantry sector⁶

The terrain here was fairly flat (fig. 171) covered with second growth and provided fair ground for tank maneuvers. The action here was characterized by temporary withdrawals of U.S. troops from forward positions under the pressure of Japanese attacks followed by highly effective tank supported counterattacks (fig. 172).

On 6 March, there were numerous patrol contacts and clashes with superior Japanese forces advancing along the Laruma River, and Allied outposts were forced back. The main attack by the Japanese 45th Infantry was launched on the morning of 12 March and succeeded in penetrating Allied wire and in occupying several pillboxes, some of which were retaken by counterattack. Again, in the early morning of 13 March, the enemy succeeded in taking six more pillboxes, and counterattacks supported by tanks resulted in retaking all but two pillboxes (fig. 173). The Japanese were attacking very strong positions in relatively open terrain, and their losses were heavy, estimated at 350-500 dead on this day, compared with 2 killed and 10 wounded in the 129th Infantry sector. The next day, 14 March, was a relatively quiet day during which Allied wire was repaired under cover of the tanks. On 15 March at 0400 hours, the Japanese again attacked and, after heavy fighting, penetrated to a depth of 100 yards over a 1,000-yard front. A tank-supported counterattack failed to dislodge the enemy who had now brought in at least one 77 mm. field gun. A second counterattack, supported by tanks and by a heavy concentration of artillery, reestablished the Allied line. Spasmodic fire occurred

⁶See footnote 5, p. 299.

305

FIGURE 171.-A cleared field of fire in front of the 129th Infantry sector.

FIGURE 172.-Light tank of the 754th Tank Battalion. This tank was in action against the Japanese at Company G, 129th Infantry, 37th Division perimeter. The cleared area in front of perimeter greatly facilitated the use of tanks.

306

FIGURE 173.-Soldiers of Company F, 129th Infantry, 37th Division, crawling up to barbed wire. Japanese were just in front of and to left of the wire and occupied the American pillboxes to the left and to the right (not shown in picture). American troops were surrounded until tanks were called upon to knock out the enemy.

on 16 March, but on 17 March at 0400 hours the enemy again attacked, breaching Allied wire to a depth of 75 yards where the attacks stopped and the enemy dug in. Prisoners' statements indicated that Allied artillery had taken a huge toll in the support and reserve units. Allied artillery continued a heavy harassing fire (fig. 174), and except for sporadic fire fights the sector was relatively quiet until 24 March when shortly after midnight the Japanese began to infiltrate. By daylight, the enemy had penetrated 300 yards (fig. 175). During the day, there was heavy hole-to-hole fighting and tank-supported counterattacks (fig. 176) which regained control of the high ground. During the latter fighting, the Japanese losses were large (fig. 177), 310 dead were counted within Allied wire compared to U.S. losses of 16 killed and 42 wounded. The artillery placed an extremely heavy concentration in front of Allied lines (fig. 178) following which only sporadic attempts to penetrate Allied wire occurred.

Table 61 gives a breakdown of the 450 casualties that were sustained by the 1,850 U.S. troops engaged on the 129th Infantry sector.

Approximately 4,300 Japanese troops were engaged on the 129th sector up to 16-17 March when an additional 600 men were brought into the area. The actual count of enemy dead was 2,373. The ratio of U.S. dead (KIA plus DOW) to Japanese dead was 1 to 30.

307

FIGURE 174.-An area devastated by U.S. artillery shell fire.

FIGURE 175.-Scene at area command post during action of Japanese infiltration of 2d Battalion, 129th Infantry, 37th Division. Reinforcing troops of Company A are in prone positions as an enemy machinegun opens fire. Tanks in background were called upon to knock out enemy positions. Burning jeep is the result of a Japanese grenade.

308

FIGURE 176.-Scene of a General Sherman medium tank and infantrymen attacking Japanese positions along the perimeter of 129th Infantry, 37th Division.

FIGURE 177.-Japanese killed on the perimeter of Company F, 129th Infantry, 37th Division. The enemy dead were hit by so many missiles it was impossible to determine cause of death.

309

FIGURE 178.-Japanese foxholes under bank of draw in 129th Infantry sector. Note how jungle was cleared by artillery fire.

TABLE 61.-Distribution of 450 casualties among 1,850 U.S. troops engaged on 129th Infantry sector, by category

Comment on Relatively Large Number of Japanese Casualties

Since one of the purposes of this study was to make observations on the relative lethal effects of weapons, the great disproportion between enemy and U.S. casualties deserves some comment. It is estimated that the enemy had 8,527 killed in action out of 10,000 troops involved in combat, as contrasted

310

to 210 killed (180 KIA plus 30 DOW) out of 5,800 U.S. troops involved. This is a ratio of 23.9 Japanese for each 1 of U.S. forces killed.

The approximate time for the Japanese attack was known, as well as the most likely points of attack. Consequently, the enemy attacked against extremely well prepared positions. United States supplies of ammunition were abundant and easily accessible to the front by an excellent system of roads. The concentration of firepower, especially artillery and mortar, was intensive. United States artillery concentration on Japanese reinforcements moving over restricted jungle tracks was particularly effective. United States forces had complete control of the air making it easy to observe, as well as to bomb, enemy troop concentrations. The limited supply of enemy artillery and ammunition had to be transported under great difficulties over the most rugged terrain. Furthermore, as in other campaigns in the South Pacific, enemy artillery was never used in concentration as judged by U.S. standards. Whenever the Japanese broke through Allied lines, which they did repeatedly, they never appeared to have sufficient reserves to follow up the advantage. There is evidence that the concentration of U.S. artillery fire on Japanese reinforcements prevented the accumulation of any effective body of troops.

On Hill 260, the ratio of Japanese dead to U.S. dead was 8 to 1. This was the most favorable ratio for the Japanese in any of the three sectors. The enemy had taken the hill very early and acquired the advantage of the terrain. Consequently, the action of U.S. troops was mostly offensive under the disadvantage of retaking a hill in which the enemy occupied well dug-in positions. The terrain prevented the use of tanks, and the proximity of the lines limited the use of U.S. artillery.

On Hill 700, the ratio of Japanese dead to U.S. dead was 36.3 to 1. While it was necessary here, also, to retake the crest of the hill, the major part of U.S. action was defensive in well-prepared positions. The enemy approach to this sector was limited because of the terrain, making artillery concentrations on their reinforcements highly effective and accounting for the greater number of enemy dead. On reaching the vicinity of U.S. lines, the enemy attacked up steep slopes in great concentration.

On the 129th sector, the ratio of Japanese dead to U.S. dead was 30 to 1. Here the approach for the enemy via the Numa-Numa Trail was easier, and the terrain permitted attack on a wider front. The terrain was also favorable to the use of tanks, and these were highly effective in retaking positions lost after the enemy had exhausted the force of their initial impact and their reserves had been disrupted by U.S. artillery. On this sector, also, the enemy attacked in great concentration on a narrow front against strongly prepared positions. Against these concentrated attacks, the use of canister-type ammunition was highly effective.

Control of the air, the use of tanks, and superior firepower in defensive positions, in addition to the greater and more effective concentrations of artillery fire, were the chief factors accounting for the large number of the enemy dead.

311

DISPOSITION OF BATTLE CASUALTIES AND ANATOMIC DISTRIBUTION OF WOUNDS

In this chapter, the term "battle casualty" is used to designate only those combatants who were killed or wounded by weapons. All deaths or injuries produced by other agents, such as falling trees, motor vehicle accidents, or others of a similar nature have been excluded. The total number of casualties includes all those wounded both by Allied and enemy weapons. Wounds caused by Japanese weapons and those resulting from U.S. weapons have been separated and are discussed under separate sections. It was impossible to ascertain which of the self-inflicted wounds were due to the soldiers' willful misconduct and which were accidental. These wounds are included and discussed in the section on U.S. weapons. It is known that 12.3 percent of the total casualties were produced by U.S. weapons. The actual percentage, however, may be slightly greater, for it is known that the enemy did use some U.S. captured weapons, particularly rifles and grenades.

There were 2,335 battle casualties. Of these, 547 (23.4 percent) were lightly wounded and were returned to duty immediately from the battalion aid or collecting stations. These 547 casualties are included in the initial total for the sake of completeness, for it was assumed that reports of casualty studies in other armies are based on computations which also include this group of minor wounds. However, in the remainder of the study, these patients have been excluded, because of the insignificant disability entailed by their injuries. Therefore, this study was based primarily on 1,788 casualties who were killed in action or who sustained wounds which necessitated hospital treatment. The term "hospital" includes two augmented clearing stations. The majority of patients returned to duty in the first echelon⁷ were treated in these clearing stations. With few exceptions, all patients who were returned to duty in the first echelon left the hospital within 30 days. Those patients in a hospital of the rear echelon, who were not evacuated to the United States, were usually returned to duty within 120 days.

Since no exact definition for the term "killed in action"⁸ has been established, an arbitrary standard was selected. In this study, KIA (killed in action) includes only those killed instantly, those found dead, and those who were mortally wounded and died shortly thereafter. Reports from division surgeons invariably contained a greater number of KIA than are found in this study. Explanation for this discrepancy is apparent and lies in the fact that the battalion surgeon frequently included, among the KIA's, patients who were initially seen alive but who were known to have died later. In this particular campaign, because of the close proximity of hospitals to the front, a large number

⁷In this chapter, "first echelon" is defined as the beachhead perimeter on Bougainville Island.-J. C. B.
⁸The usual definition is: Wounds directly due or attributable to enemy action which result in death before the casualty is admitted to a medical installation or receives treatment from a medical officer.-J. C. B.

312

of casualties are included under WIA (wounded in action),⁹who perforce, under less favorable circumstances, would have been classified as killed in action.

Table 62 gives a breakdown of the Bougainville casualties during the survey period and the general disposition of the WIA. It may be seen that the 395 dead (320 KIA and 75 DOW) constitute 16.9 percent of the total casualties. Thus, there was approximately one battle death (KIA plus DOW) among every six casualties (WIA, including DOW, plus KIA). Nearly 70 percent of all casualties were returned to duty within the theater and, of the 1,940 living wounded, 1,622 (83.6 percent) were returned to duty. However, 547 of these were returned to duty from a first aid post and did not require hospitalization. These soldiers had very minor wounds and were not lost to battle. Since the incapacitating effect of weapons on this group was negligible, they were eliminated from the remainder of this study, leaving 1,788 casualties who were killed or whose wounds were of such severe degree that they were lost to the battle. Using this criterion, there was approximately one battle death (KIA plus DOW) among every four and a half casualties (WIA, including DOW, plus KIA). The WIA (including DOW)/KIA ratio was 4.6:1. Those who died and those who were evacuated to the United States were classed as "lost to service" and comprised 30.5 percent of the total casualties.

A study of both the living and the dead is essential in order to gain an accurate and complete picture of the anatomic distribution of wounds produced

TABLE 62.-Distribution of 2,335 Allied casualties in Bougainville campaign, from 15 February to 21 April 1944, inclusive, by category

⁹The usual definition is: Wounds directly due or attributable to enemy action which necessitate admission of the casualty to a medical installation and treatment by a medical officer. This generally includes those who are wounded treated, and died later (died-of-wounds (DOW)) or preferably wounded-treated-died (WTD).-J. C. B.

313

by various weapons. Although many wound studies have been made on the living, few records are available which analyze the effect of weapons on both the dead and living. In this investigation, data concerning all those who were killed in action as well as those who were wounded in action and died later have been collected and combined with the records of the living wounded.

Information regarding the circumstances of wounding in the living is relatively easy to obtain. Frequently, the facts may be elicited by an interview with the person wounded. However, the information will be still more accurate if checked with an eyewitness. To secure accurate details concerning the dead, however, is much more difficult. Post mortem examinations should be done, of course, whenever possible. Autopsies, however, were limited by the fact that all bodies could not be recovered and also by the fact that some were decomposed when recovered. Unfortunately, rapid deterioration occurs in the tropical climate of Bougainville, and for sanitary reasons the dead must be buried as soon as possible. The dead, when recovered, frequently exhibit wounds other than those which produced death. Wounds inflicted after death were especially common in areas subjected to concentrated artillery or mortar fire. Furthermore, it was often difficult and frequently impossible to identify the lethal weapon from the appearance of the wound or the missiles recovered at autopsy. In many instances, discrepancies were found when the emergency medical tag, hospital record, and post mortem findings were compared. It became apparent, therefore, that the true sequence of events leading to death could be secured only by careful personal questioning of witnesses who saw the soldier killed or who knew personally of the circumstances surrounding his death. By adhering to this method of investigation, a relatively high degree of accuracy was achieved, not only in the records of the dead but also of the living.

All casualties were classified under anatomic regions according to the location of the wound. In many instances, a major wound was accompanied by one or more minor wounds. In this event, the anatomic location of the major wound alone determined the classification. Furthermore, if a single wound among others was responsible for death or disability, the anatomic location of that wound determined the classification. In the classification of the casualties, it became necessary to add to the conventional division by anatomic regions, an additional group which was designated "Multiple Wounds." The term "multiple wounds" is used for those casualties sustaining two or more wounds, either one of which might have been responsible for the soldier's death or for rendering him unfit for action. It was difficult or impossible to classify accurately all casualties who received more than one wound. In many instances, the dead were struck by other missiles after death, under which circumstances it was not possible to decide which of several wounds produced death. In other instances, decomposition of the body made examination unsatisfactory. For these and other reasons, some patients were placed in the multiple wound classification, who probably should have been included properly with those grouped under single anatomic regions. Because of the diffi-

314

culty in analysis of the "multiple wounded," every effort was made to keep to a minimum the number so classified. Nevertheless, the "multiple wounded" constituted 18.6 percent of the casualties (table 64).

It is desirable in a study of this kind, if possible, to evaluate the influence of various factors on the anatomic distribution of wounds. Particular consideration should be given to the type of action (defensive or offensive), available cover or protection, armor, terrain, and type of weapon and projectile employed. Furthermore, if a true representation of the distribution of wounds is to be established, the data should be derived from a study of the dead as well as the living.

In table 63, the anatomic distribution of wounds in the living and dead in Bougainville is compared with similar wound distributions in the living in two past wars and in World War II. It will be observed that head wounds were more frequent at Bougainville than elsewhere. Perhaps this was due to the relatively close range of rifle fire in jungle warfare. Another discrepancy is observed by comparing the percentage of wounds of the head, chest, and upper extremities in the living. For example, wounds of these regions on Bougainville attained a total of 60.5 percent, whereas the Russians in World War II reported a total of only 48.5 percent.

TABLE 63.-Comparison of wounds in living wounded of two past wars and World War II with casualties of Bougainville campaign, 15 February to 21 April 1944, inclusive, by anatomic location

¹Multiple wounds excluded.
Source: Monthly Progress Report, Army Service Forces, War Department, April 1944, Section 7: Health.

With the exception of head wounds, the anatomic distribution of wounds in jungle warfare does not appear to differ greatly from the distribution of wounds reported for other types of warfare. In the absence of available data on other types of warfare, it is difficult to derive an adequate explanation for this high frequency of head wounds. The mortar followed closely by the rifle was the most frequent cause of wounds of the head. Since rifles are used frequently, and at close range, in jungle warfare, it is suggested that the greater

315

number and accuracy of bullets might account for the high incidence of head wounds. However, no proof can be offered for this hypothesis. The factor of exposure appears to offer no better explanation, since the head is apparently exposed to the same degree in jungle as in other types of warfare. The predominance of lower extremity wounds is accounted for by the high incidence of mortar hits.

Table 64 shows the anatomic distribution (regional frequency) of wounds in the 1,788 Bougainville battle casualties.

TABLE 64.-Distribution of wounds in 1,788 battle casualties, by anatomic location (regional frequency)¹

¹Any casualty with major wounds in more than one anatomic region is cataloged under "Multiple." Therefore total number of wounds is same as total number of casualties.
²Percent for dichotomy, dead versus living, by each anatomic location and for total dead versus living.

The anatomic distribution of wounds in the dead (table 65) is in striking contrast to the distribution of wounds in the living (table 66). The low incidence of extremity wounds among the dead is a rough index of the effectiveness of modern surgery, when dealing with wounds which do not involve a vital organ. Multiple wounds hold second place among the dead. Bullets (rifle and machinegun), mostly at close range, caused 58.2 percent of all deaths (table 77), while high explosives (artillery, mortar, grenade, and mine) caused 38.8 percent. A consideration of those who were wounded in action and died later (table 65) indicates that the major problem is encountered in wounds of the abdomen and thorax. These two regions accounted for 65.3 percent of all those who were wounded in action and died later.

An index of the degree of the residual disability may be obtained by a consideration of the number of patients returned to duty or evacuated to the United States (table 66). It should be noted that while the total number of patients in the anatomic divisions varies considerably, the percentage of patients returned to duty in each anatomic region remains remarkably constant.

316

TABLE 65.-Distribution of wounds in 395 dead, by anatomic location

¹Percent killed in action or died of wounds by each anatomic location of total killed in action or died of wounds, respectively.
²Percent of dichotomy, killed in action versus died of wounds, by each anatomic location and for total killed in action versus died of wounds.

TABLE 66.- Distribution of wounds in 1,393 living wounded, by anatomic location

¹Percent for dichotomy, returned to duty versus evacuated to United States, by each anatomic location and for total returned to duty versus evacuated to United States.

Table 67 lists the regional frequency of wounds and the disposition of the living wounded in the 1,788 casualties.

Head wounds alone were responsible for 384 or 21.5 percent of all battle casualties. Of the 134 KIA in this group, death resulted from brain injury in 125 and from wounds of the face and neck in 9. In the 10 patients who were wounded in action and died later, 9 sustained brain injuries and 1 a transection of the cervical spinal cord. A more detailed description of these 10 patients will be found in another section under "Treatment of the Wounded." Of the surviving 240 patients, 157 (65.4 percent) were returned to duty in the first echelon.

317

TABLE 67.- Distribution of 1,788 battle casualties, by disposition and anatomic location of wounds (regional frequency)

¹Defined as the beachhead perimeter on Bougainville Island.
²From hospitals on Guadalcanal, Espíritu Santo, and New Caledonia.

318 

In table 68, the head wounds previously summarized (in table 67) are combined with those head wounds which are described later under "Multiple Wounds," making a total of 505. It is evident by comparison of the two tables that the ratio of the dead to those evacuated to the United States, and to those returned to duty, remains relatively unchanged. The inclusion of multiple wounds with those classified under single anatomic regions may lead to duplication and confusion. For this reason, multiple wounds have not been included in any tables except those devoted to the analysis of head wounds.

TABLE 68.-Distribution of 505 casualties with head wounds (including multiple wounds), by category

¹Defined as the beachhead perimeter on Bougainville Island.
²From hospitals on Guadalcanal, Espíritu Santo, and New Caledonia.

Thoracic wounds accounted for 12.9 percent of all battle casualties. Of the dead, 66 were killed in action. Of the 21 who were wounded in action and died later, 15 died during or following operation. Perforating wounds of the thorax were present in all those who were killed or died later. Of 63 patients returned to duty in the first echelon, only 3 had wounds which penetrated the pleural cavity; all others had wounds of the chest wall only. In the group of 47 patients returned to duty from the rear echelon, 33 sustained chest wall wounds only. Among the remaining 14 with lesions involving the lung or pleura, 6 underwent lung operation. Of 34 patients who were evacuated to the United States, 24 had injuries of the lung; 19 of this latter group were treated by surgical operation and 5 by conservative measures. The remaining 10 patients had wounds of the chest wall which did not communicate with the pleural cavity.

319

A total of 114 patients sustained abdominal wounds. The abdomen was struck less frequently than any other anatomic region, and these wounded constituted the smallest number (6.4 percent) of all casualties. In 10 patients, wounds involving both the abdomen and thorax with perforation of the diaphragm were present. Twenty patients were killed in action, one of whom sustained a transection of the spinal cord. A relatively greater number (28, 24.6 percent) of patients were wounded in action and died later in this group than any other. Of these 28 patients, only 3 died without operation. In most instances, death resulted either from shock and hemorrhage or from peritonitis. The entire group of 19 patients returned to duty in the first echelon had wounds of the abdominal wall only. Of 23 patients returned to duty from the rear echelon, 13 had abdominal wall wounds; 1 had a combined thoracoabdominal wound; and the remainder had visceral lesions distributed as follows: Liver, 4; colon, 2; spleen, kidney, and bladder, 1 each. Fewer patients wounded in the abdomen (36.9 percent) were able to return to duty, than were those wounded in any other region. Of the 24 patients evacuated to the United States, 18 had injuries of the abdominal viscera, 5 had abdominal wall wounds, and 1 a transection of the cauda equina. The visceral lesions among these patients were distributed as follows: Small intestine, 6; small intestine and colon, 4; colon, 3; spleen and diaphragm, 2; stomach and liver, colon and diaphragm, and bladder, 1 each.

Wounds of the upper extremity alone constituted 17.9 percent of all battle casualties, yet wounds of this region carry a death risk of only 0.3 percent. No patients died who received treatment. The number of patients returned to duty in the first echelon is greater among those receiving upper extremity wounds than among those wounded in any other region. Of these 175 patients, 4 had fractures of the hand and 2, incomplete fractures of the arm. In the 68 patients returned to duty from the rear echelon, there were 12 fractures as follows: 6 of the bones of the hand and 2 each of the scapula, humerus, and forearm. In the 76 patients evacuated to the United States, there were 58 compound fractures and 5 amputations. The fractures were distributed as follows: Humerus, 23; bones of the forearm, 19; bones of the hand, 12; and scapula, 4. The percentage of patients evacuated to the United States was higher in upper extremity wounds than in wounds of any other anatomic region.

Wounds of the lower extremity were the most numerous of all battle wounds (22.7 percent) and accounted for next to the lowest mortality of any region (1.5 percent KIA). There were six casualties classed as killed in action, although with one exception all were alive when first seen. These soldiers either could not be reached or else died before adequate medical aid could be given. There were eight patients who were wounded in action and died later. Seven of these died in the first echelon and one in the second echelon. Of the seven deaths in the first echelon, two resulted from gas gangrene and five from shock and hemorrhage. Two deaths in the latter group might have been avoided by the use of a tourniquet. From the first echelon, 195 patients were

320

returned to duty. With the exception of one patient who had a chip fracture of the tibia, all of these patients had soft-tissue wounds only. From the rear echelon, 113 patients were returned to duty, 8 of whom had fractures of the bones of the leg and 4 of the bones of the foot. In 85 patients evacuated to the United States, there were 58 compound fractures distributed as follows: Bones of the leg, 31; femur, 18; and bones of the foot, 9. In addition, there were 10 amputations of the thigh or leg.

The risk of death in wounds of the extremities is low. In 727 casualties with wounds of the upper and lower extremities, there were 15 deaths (2.0 percent). On the other hand, wounds of the extremities constituted half of all patients evacuated to the United States. The majority of patients with wounds of the extremities, who were lost to the service by evacuation, had fractures as shown in table 69. Fractures among upper extremity wounds are more common (29.5 percent) than among the lower extremity lesions (18.3 percent). The greater relative volume of soft tissue to bone in the lower extremity may explain the lower incidence of fracture. On the other hand, the explanation may lie in the fact that the percentage of high-velocity missile

TABLE 69.-Disposition of 319 casualties with wounds of upper extremities and 393 casualties with wounds of lower extremities

¹Percent for dichotomy, fractures versus nonfractures, under each disposition category and for total fractures versus nonfractures by upper and lower extremity wounds.
²Defined as the beachhead perimeter on Bougainville Island.
³From hospitals on Guadalcanal, Espíritu Santo, and New Caledonia.

321

wounds are slightly greater in the upper than in the lower extremity. Bullets produced 36.9 percent of all wounds of the upper extremity and 27.9 percent of the wounds of the lower extremity. Patients who returned to duty in the first and second echelons usually had fractures of small bones, chip and perforating fractures, and other fractures with minimal bone damage. It should be noted that 89.5 percent of the patients with wounds of the upper extremity and 68.2 percent of those with wounds of the lower extremity were evacuated to the United States because of fractures. The cause of fractures is discussed further in a later section devoted to the relative effect of weapons.

Wounds were classed as multiple only if two or more wounds of different regions could have caused death or disability. Such wounds caused 18.6 percent of all battle casualties. As in wounds of the head when death occurred, it was usually instantaneous. On the other hand, a relatively high percentage of patients with multiple wounds were returned to duty. In a group of 91 patients returned to duty from the first echelon, there were 203 soft-tissue wounds distributed as follows: Upper extremity, 73; lower extremity, 61; thoracic wall, 22; face and neck, 20; scalp, 14; abdominal wall, 8; and eye, 5. Present also were chip fractures of the clavicle, finger, and leg. From the rear echelon, 82 patients were returned to duty with 186 soft-tissue wounds distributed as follows: Upper extremity, 62; lower extremity, 61; thoracic wall, 24; face and neck, 17; scalp, 10; abdominal wall, 8; abdominal perforations, 2 (spleen and rectum); eye, 1; and lung perforation, 1. There were 8 chip fractures, 6 of the upper and 2 of the lower extremity, and also 2 perforating fractures of the pelvis; in addition, there were 2 finger amputations. There was a total of 151 soft-tissue wounds and fractures in 58 patients who were evacuated to the United States. The 38 fractures were distributed as follows: Upper extremity, 20; lower extremity, 16; and jaw, 2. The following soft-tissue wounds were present: Upper extremity, 36; lower extremity, 34; thoracic wall, 16; face and neck, 12; eye, 5; amputations, 5; scalp, 3; brain, 1; and abdominal wall, 1. Among these patients with multiple wounds, fractures were the chief cause for evacuation to the United States.

The anatomic distribution of wounds may vary according to the type of weapon causing the wound, the degree of exposure of different parts of the body, the protection afforded by various means, and the direction of fire. If the body were unprotected in an atmosphere of flying missiles of equal distribution, wounding should occur in direct proportion to the exposed surface area. However, such a theoretical condition never exists. On the contrary, missiles usually move in one direction at a given time. The projected area of the body if completely exposed, therefore, offers a better measure for the study of the probable hits. The mean projected body area¹⁰ is obtained from projection in three positions, standing, kneeling, and lying. The hits with all weapons are compared with the mean projected body area (table 70). The head is the only region in which the percentage of wounds appreciably ex-

¹⁰Burns, B. D., and Zuckerman, S.: The Wounding Power of Small Bomb and Shell Fragments. R. C. No. 350 of the Research and Experiments Department of the Ministry of Home Security.

322

ceeded the percentage of the projected area for that region. The percentage of hits in the abdominal area is considerably less than the percentage of its projected area. The question may be raised why wounds of the head so far exceed the projected head area. Was this due to good marksmanship or exposure? Obviously, the head must be exposed for marksmanship to be effective. Since wounds caused by rifle bullets and mortar shell fragments were found in significant numbers and the circumstances were known with reasonable accuracy, they may be compared. The directed fire of the rifle and the undirected hits with mortar fragments were found to approximate closely the total hits by all weapons. This is evidence that exposure is one of the chief factors in accounting for the high incidence of head wounds. Nevertheless, the number of wounds caused by rifle fire does exceed the number caused by mortar fragments in the head, upper extremity, and thorax. This may be interpreted as evidence that marksmanship does play a small but important part in the high incidence of head wounds. This observation is further substantiated by the fact that the lower extremity presents the reverse of these findings.

TABLE 70.-Mean projected body area and wound distribution (excluding multiple wounds)

¹Includes all other weapons in addition to rifle and mortar which are shown specifically.

THE DIFFERENT WEAPONS CAUSING BATTLE CASUALTIES

It is obvious that the number of battle casualties produced by various weapons will depend upon the type of warfare, the number of weapons employed, and the training and tactics of the opposing forces. Thus, the measure of effectiveness of a given weapon must, of necessity, vary according to the circumstances under which it is used. The effectiveness of a weapon depends not only upon the total number of casualties it produces but also upon the ratio of the killed to wounded and upon the severity of the wound. In a certain local situation, the most effective weapon might be one which temporarily disabled the greatest number of the enemy and hence allowed the capture of a particular

323

objective or the winning of a single battle. If the effectiveness of a weapon is to be measured by this latter criterion, it would be necessary to set up an arbitrary definition of "temporary disability." In this event, a solution of the problem would be found in classifying the wounded on the basis of "ability to continue combat if life depended upon it."

The ratio of the killed to wounded is subject to various interpretations and must be clarified. As previously stated, the term "killed in action" in this study indicates those killed instantly and those who were mortally wounded and died within a relatively short time. Because of the proximity of medical installations on Bougainville, many mortally wounded patients lived to reach the hospital and were classified among those who were wounded in action and died later. Doubtless, under other less propitious circumstances, many of these casualties would have been classified with those who were killed in action. The term "dead" refers to the total number of those killed in action and those who were wounded-treated-died-later.

Since the severity of a wound is an abstract quality, open to individual interpretation and judgment and hence to consequent error, it was necessary to establish another criterion by which to judge the degree of disability sustained. The ultimate disposition of the patient seemed to offer a more reasonable basis for this estimation. All wounded, therefore, were separated into three groups depending upon whether the nature of the wound allowed the patient to be returned to duty from the first or from the rear echelon or whether it necessitated his evacuation to the United States. It is recognized that this is an arbitrary standard and open to the criticism that it is also an index of medical care; nevertheless, it is a factual and objective measure of the relative effect of weapons in the living wounded.

A fairly comprehensive description of the common types of Japanese weapons used on Bougainville has already been presented (pp. 289-292). From wound examination alone, it was never possible to distinguish the caliber of rifle or machinegun bullets nor the size of explosive shells. It was frequently impossible to judge with any accuracy whether the wound had been produced by a bullet or grenade shell or bomb fragment. Aerial bombing by the enemy did not occur during the Battle of the Perimeter. Miscellaneous weapons producing wounds were the bomb (U.S. aerial bombs), 13; pistol, 13; bangalore torpedo, 9; powder explosion, 5; bayonet, 2; bazooka, 1; and parachute flare, 1.

The Relative Lethal Effect of Weapons

The phrase "relative lethal effect" of a weapon refers to the percentage of deaths among the total number of casualties (dead and wounded) caused by that particular weapon. As previously stated, the ratio of the number of deaths to the number of casualties produced by any given weapon depends upon such variable factors as the type of action (offensive or defensive), number of weapons employed, terrain, exposure, and available protection. These factors

324

determine primarily the necessary degree of exposure of the soldier and consequently the number of hits, other factors being equal. The type and number of the particular weapon employed is then of prime importance in determining the relative lethal effect. For example, a small number of machineguns may produce few casualties but a "high lethal effect,"¹¹ whereas a great many casualties may result from heavy mortar fire yet the lethal effect will remain relatively low.¹²

A comparison of the incidence of casualties caused by different weapons (table 71) shows that the mortar wounded more men (38.8 percent) than any other weapon. This was the weapon most feared by Allied troops. However, the relative lethal effect of the mortar is low (11.8 percent), rating next to the grenade which has the lowest (6.2 percent) relative lethal effect. There were 1,741 casualties caused by HE shells, grenades, landmines, and bullets and 47 casualties produced by miscellaneous weapons. High explosive shells, grenades, and mines caused wounds in 1,145 men (64.1 percent), but only 153 deaths (38.7 percent) occurred in this group. In contrast, bullets hit a total of 596 men (33.3 percent), but they accounted for 230 deaths (58.3 percent of total hit). The rifle was responsible for wounds in 445 casualties with a lethal effect of 32.1 percent. The machinegun, while causing fewer casualties (151), had the highest lethal effect of 57.6 percent. The very low lethal effect of the grenade (6.2 percent) is a characteristic probably peculiar to the Japanese hand grenade. Of the 34 landmine casualties, 33 were produced by U.S. mines. The 47 casualties (2.6 percent) listed under miscellaneous weapons were caused by pistols, bangalore torpedoes, bazookas, flares, powder explosions, and bayonet wounds.

Table 72 is a breakdown of the various causative agents according to the anatomic distribution (regional frequency) of wounds in the 1,788 casualties.

TABLE 71.-Distribution of 1,788 battle casualties, by relative lethal effect of causative agent

¹Percent for dichotomy, dead versus living, by each causative agent and for total dead versus living.

¹¹High mortality-low morbidity.-J.C.B.
¹²High morbidity-low mortality.-J.C.B.

325-326

There were 384 casualties (21.4 percent of the total number) due to wounds of the head alone. Moreover, wounds of the head (144) accounted for 37.5 percent of all dead. Excluding the 5 wounded by miscellaneous weapons, 208 head casualties (54.2 percent) were produced by high explosives (fragments) and 171 (44.5 percent) by bullets. However, high explosives accounted for

TABLE 72.-Relative lethal effect of weapons, by anatomic location of wounds and for multiple wounds

¹Percent for dichotomy, dead versus living, by each causative agent and for total dead versus living by anatomic location of wounds and for multiple wounds.

only 27.1 percent of the dead, whereas bullets were responsible for 72.9 percent. Thus, while high explosives caused more casualties, the lethal effect produced was relatively low. This may be explained by the average lower velocity of shell fragments and the relative greater protection afforded against them by

327

the helmet and skull. This is further substantiated by the fact that in 92.3 percent of the deaths due to head wounds, the skull had been penetrated.

Wounds of the thorax accounted for 12.9 percent of all casualties and for 22.0 percent of all deaths. Excluding 5 wounded by miscellaneous weapons, high explosives (fragments) produced 135 casualties (60 percent) and bullets 91 (40 percent). However, again contrasting relative lethal effects, bullets accounted for 59.7 percent of the deaths and high explosives for 40.3 percent. In thoracic wounds, the contrast between the lethal effect of wounds due to high explosives and bullets is not so pronounced as in wounds of the head. Possibly, this is due to the fact that the thoracic cage offers less protection to the vital organs than does the skull and helmet. This hypothesis seems to be substantiated further by the fact that while the lethal effect of both mortar and artillery fragments is increased in the thorax, the lethal effect of the grenade is increased fourfold. Bullet wounds were limited to the chest wall in only 18 instances, while high explosives caused 85 wounds which did not penetrate the thoracic cavity. The relatively lower velocity of some of the HE shell fragments would appear to account for its frequent failure to penetrate the thorax.

Casualties occasioned by wounds of the abdomen had the lowest incidence and accounted for only 6.4 percent of the total wounded and 12.1 percent of the dead. Whereas, high explosives (fragments) caused 56.2 percent of the casualties due to abdominal wounds, bullets accounted for 62.5 percent of the deaths from these wounds. This ratio may represent a distorted picture when compared to findings in other theaters, since it is based on such a small number (8) of wounds of the abdomen caused by artillery shells. However, the mortar and the grenade show almost twice the relative lethal effect in wounds of the abdomen as they do in wounds of the thorax. This is further evidence that the bony structures of the body wall may offer considerable effective protection against these low-velocity fragments. High explosive fragments caused 30 of the 53 wounds perforating the abdominal cavity, which would appear to indicate a relatively high index of penetration. Nevertheless, the relative protection afforded by the abdominal wall to low-velocity fragments should also be mentioned. Of 38 wounds limited to the abdominal wall, 30 were caused by HE fragments.

Wounds of the upper extremity accounted for 17.9 percent of all casualties and for only 0.3 percent of the dead. High explosive fragments caused 59.1 percent of these wounds. More than half of all wounds caused by high explosives were due to mortar shells. The relative effectiveness of bullets and HE fragments may be judged from the severity of the wound as indicated by the disposition of the patients shown in table 73.

The one death among the upper extremity casualties was caused by a mortar shell. Since the lethal effect of wounds of the upper extremity was negligible, it deserves no discussion.

Wounds of the lower extremity accounted for the highest number of casualties (22.7 percent). However, lower extremity wounds were responsible for only 3.5 percent of all deaths. High explosives caused 70.8 percent of lower

328

extremity casualties; of these, mortar shells alone were responsible for more than half. Bullets, however, caused 7 of the 14 deaths. The severity of wounds caused by bullets and high explosives may be judged by the disposition of casualties as shown in table 74.

TABLE 73.-Disposition of 123 and 196 casualties with upper extremity wounds, by relative effectiveness of bullets and HE fragments, respectively

¹Defined as the beachhead perimeter on Bougainville Island.
²From hospitals on Guadalcanal, Espíritu Santo, and New Caledonia.

TABLE 74.-Disposition of 110 and 283 casualties with lower extremity wounds, by relative effectiveness of bullets and HE fragments, respectively

¹Defined as the beachhead perimeter on Bougainville Island.
²From hospitals on Guadalcanal, Espíritu Santo, and New Caledonia.

Wounds of the extremities constituted the largest group of battle casualties in this survey and accounted for 40.6 percent of all wounds. These wounds, however, accounted for the smallest number of dead (3.8 percent). Since relatively few deaths resulted from wounds of this region, the effectiveness of weapons on the extremities must be judged by the duration of the soldiers' incapacity and by the number of casualties lost to the service by evacuation to the rear echelon and to the United States. In view of the fact that fractures were the chief cause of evacuation to the United States, the

329

relative effect of weapons on the extremities was also judged by the number of fractures they caused. The rifle caused the greatest number of fractures in both the upper and lower extremities (table 75). In the upper extremity, the rifle led not only in the number but also in the percentage chance of fracture. In general, the chance of fracture appeared to parallel the velocity of the missile. Bullets caused only 37.5 percent of upper extremity and 26.3 percent of lower extremity wounds, whereas these missiles caused 66 percent of upper extremity and 60 percent of lower extremity fractures.

TABLE 75.-Relative effect of weapons causing wounds of upper and lower extremities, among the living wounded

¹Percent for dichotomy, fracture versus nonfracture, by each causative agent and for total fracture versus nonfracture, by upper and lower extremity wounds.

Casualties due to multiple wounds rated third in incidence and constituted 18.6 percent of the total number. High explosives caused 79.5 percent of these wounds and 53.5 percent of the resultant deaths; however, the machinegun and rifle showed the highest relative lethal effect. The severity of multiple wounds caused by bullets and high explosives as judged by the disposition of casualties is shown in table 76.

330

TABLE 76.-Disposition of 16 and 215 casualties with multiple wounds, by relative effectiveness of bullets and HE fragments, respectively

¹Defined as the beachhead perimeter on Bougainville Island.
²From hospitals on Guadalcanal, Espíritu Santo, and New Caledonia.

The Dead

Table 77 shows the distribution of the dead according to the causative weapon. There were 395 dead of whom 230 or 58.2 percent were killed by bullets. Of these 395 dead, 75 (19 percent) were wounded in action, treated, and died later. Of these 75 patients, 50 died within 24 hours; of these 50, 40 were classed as mortally wounded. Had medical facilities been further removed from the frontline or had transportation problems been more difficult, a large number of those who were wounded and died later would, no doubt, have been classed as KIA. Bullet wounds tended to produce more immediate fatalities than did wounds produced by mortar and artillery shells. Among those who were wounded and died later, wounds were produced by the mortar in 28.0 percent, by artillery in 27.3 percent, and by the rifle in 14.7 percent.

TABLE 77.-Distribution of 395 fatal casualties, by relative effect of causative agent

¹Percent for dichotomy, killed in action versus died of wounds, by causative agent and for total killed in action versus died of wounds.

331

Effectiveness of Weapons

To measure the effectiveness of a weapon by the number of casualties it produces may lead to erroneous conclusions. To reiterate, the number of casualties depends on such factors as the necessary exposure of the soldier, the concentration of troops, the number of weapons employed, and the effect of the missile. It is seldom that all these varying conditions of battle can be duplicated. On the other hand, the percentage chance of death and the length of disability when hit by a given weapon should remain relatively constant and, therefore, should offer a fairly accurate index of the effectiveness of various missiles.

The percentage chance of death when hit by various weapons is shown in table 78. Casualties receiving two or more wounds, either one of which might have produced death, are not included in this table, but are discussed under "Multiple Wounds." Nevertheless, many of these casualties did have more than one wound. The order of these weapons suggests that the chance of being killed is a function of the velocity of the missile. The risk of death when hit by a machinegun in the head, chest, or abdomen is approximately equal. The contrast in death risk between the machinegun (54.5 percent) and the rifle (29.1 percent) is not entirely due to multiplicity of hits, since multiple hits were found not infrequently with rifle fire. On the average, machinegun fire originated at a closer range than rifle fire, 61 percent of the hits being from less than 50 yards. The chance of death when hit by a grenade (4.9 percent) is approximately half that when hit by the mortar (9.7 percent). The risk of death when hit in the abdomen by mortar or grenade is relatively greater than when hit in the head or thorax. This suggests that the helmet and skull (fig. 179) as well as the ribs may offer considerable protection against many of these relatively low-velocity fragments.

The relative effect of weapons may be judged by the percentage chance of a light wound or of a severe wound (tables 79 and 80). These tables are based on living wounded only. A light wound was defined as one which allowed return to duty in the first echelon and a severe wound as one which necessitated evacuation to the United States. There appears to be considerable difference in the severity of a wound according to the anatomic region hit, as well as to the weapon causing it. In general, high explosives (fragments) tend toward light wounds while small arms (bullets) tend toward more severe wounds.

The relative effectiveness of weapons may also be evaluated by a consideration of the total dead plus the total evacuated to the United States. Together, these may be considered as "lost to the service" (table 81), although some who were returned to the United States may serve in future campaigns. It should be noted by this criterion that wounds of the extremities and abdomen assume a far greater relative importance than when death alone is utilized as an index of weapon effectiveness.

332

TABLE 78.-Relative effect of weapons: Probability of hits resulting in death, by anatomic location of wounds (excluding multiple wounds)

¹Excluding mines and miscellaneous agents.

333

FIGURE 179.-Roentgenogram of skull showing artillery shell fragment lodged in sinus cavity. A soldier, standing in the company area, was hit by a Japanese 75 mm. artillery shell which exploded at a distance of 100 yards. A fragment of the shell penetrated the outer wall of the maxillary sinus and lodged in the sinus cavity. This is a good example of the relative protection afforded by bony structures to low-velocity fragments, even of large size. A. X-ray of skull. B. Recovered fragment.

Table 82 shows the number of patients returned to duty from the first echelon (Bougainville). Table 83 shows the total number of casualties dead and evacuated to the rear echelon and to the United States. These were lost to the Bougainville campaign. Note that the percentage effectiveness of each weapon suggests a possible correlation with the average velocity of hits.

Conditions of battle may be such that the effectiveness of a weapon can best be measured by whether the wounded soldier was able to continue fighting. Hence, the number of casualties per se is not a sufficient criterion since many of the wounded may continue to fight and hold off the enemy, at least temporarily. It is, therefore, desirable to know the number who are put out of action immediately and the number who could continue combat for a period of hours, if life depended on it. A questionnaire to determine whether an individual did or did not continue combat was found to be misleading, since conditions of battle were frequently such as to permit the soldier to seek immediate treatment. This he usually did when possible, since he had been so instructed by Medical Corps personnel. However, there were numerous instances of soldiers who were severely wounded and yet who continued to hold their position in the line until relieved. (For example: Two soldiers were holding a pillbox at night under Japanese attack. Eventually, each had a hand blown off, but with two hands between them, they cared for their wounds, manned their guns, and held off the attack until relieved at daybreak.)

334

TABLE 79.-Relative effect of weapons: Probability of causing light wounds¹

¹Based on percent of living wounded (survived less multiple wounded) returned to duty from first echelon (defined as the beachhead perimeter on Bougainville Island).

TABLE 80.-Relative effect of weapons: Probability of causing serious nonfatal wounds¹

¹Based on percent of living wounded (survived less multiple wounded) evacuated to the United States.

335

TABLE 81.-Relative effect of weapons: Lost to service in the theater ¹

¹Percent of hits (dead plus survived, excluding multiple wounded) resulting in death or evacuation to the United States.

336

TABLE 82.-Relative effect of weapons: Casualties returned to duty from first echelon¹

¹Defined as the beachhead perimeter on Bougainville Island.

TABLE 83.-Relative effect of weapons: Casualties lost to Bougainville campaign (dead or evacuated to rear echelon¹or to United States)

¹To hospitals on Guadalcanal, Espíritu Santo, and New Caledonia.

An arbitrary criterion based on the seriousness of the wound seemed justified in order to determine whether a soldier will be able to continue in battle for a number of hours, if his life were at stake. For this purpose, an arbitrary schedule was derived, and the following wounded were classed as "Lost to Combat":

1. Wounds of the head and central nervous system producing unconsciousness or paralysis.

2. Wounds of intrathoracic structures producing hemorrhage and shock.

3. Wounds of intraperitoneal structures producing hemorrhage and shock.

4. Wounds of the extremities producing fractures of long bones, severance of major blood vessels, or major traumatic amputations.

5. Extensive wounds of soft tissue producing shock.

The wounded were classified according to the criteria listed and added to the dead to determine the total lost to combat (table 84). This table again suggests that the percentage effectiveness of the weapon is a function of the average velocity of the missiles.

Callender and others have shown that the wounding power of a missile is in proportion to the cube of the velocity, the mass and other factors being equal. In this report, the percentage effectiveness of weapons as judged by

337

the chance of death, and the severity of the wound, appears to be in accord with the observation that the wounding power of a missile is chiefly a function of velocity. When hits occur, the weapons in order of effectiveness are (1) machinegun, (2) rifle, (3) artillery, (4) mortar, and (5) grenade.

TABLE 84.-Relative effect of weapons: Casualties lost to combat

The Relative Effect of Weapons on the Disposition of Patients

An evaluation of the effectiveness of each weapon may be obtained by considering both the number killed and the severity of the wound as determined by the disposition of the patient.

There were 700 casualties returned to duty from the first echelon (defined as the beachhead perimeter on Bougainville Island). These patients spent an average of 12.7 days in the hospital (table 85). However, if the requirement had existed, the majority of these men would have been available for emergency combat duty in a shorter time. Nevertheless, the problem of the lightly wounded, treated in the first echelon, is of considerable importance, both because of days lost to the service and because these casualties occupy beds which might be needed for the more seriously wounded. Wounds caused by HE shell fragments constituted the major problem in the first echelon. Wounds of the extremities and multiple wounds comprised a majority of these lesions (table 86).

The rear echelon included hospitals on Guadalcanal, Espíritu Santo, and New Caledonia; the evacuation distances ranged from 400 to 1,500 miles from Bougainville. Consequently, patients evacuated to hospitals in the rear were lost to the service insofar as the Battle of the Perimeter was concerned. Subsequently, some of these patients were returned to duty from the rear echelon and performed service in combat units, hence were not lost to the South Pacific theater. The severity of the wounds in these casualties usually justified their removal to a rear echelon for convalescence. Only a very few were evacuated because of the need for additional vacant hospital beds on Bougainville. Hence, transfer to the rear echelon may be taken as a fair measure of the severity of a soldier's wound from the standpoint of his ability to undergo combat. The wounded were usually returned to duty from the

338

TABLE 85.-Days lost by 700 casualties returned to duty from first echelon¹ hospitals, by causative agent

¹Defined as the beachhead perimeter on Bougainville Island.

TABLE 86.-Days lost by 700 casualties returned to duty from first echelon¹ hospitals, by anatomic location

¹Defined as the beachhead perimeter on Bougainville Island.

rear echelon or were evacuated to the United States within 120 days. However, the average elapsed time before return to duty was considerably less than this.

Though many patients evacuated to the United States were returned to duty eventually, they must be considered as lost to the service for a long period.

Table 87 presents the anatomic distribution of the hits by the various causative agents, and table 88 lists the general disposition of the nonfatal casualties.

In number of wounds produced, the rifle was exceeded only by the mortar and was responsible for 24.9 percent of all battle casualties. However, the rifle ranked first as a lethal agent, accounting for 36.2 percent of all dead. Moreover, it was second in percentage relative lethal effect (32.1 percent), being exceeded only by the machinegun (57.6 percent). The rifle produced

339-340

wounding in 53.7 percent of all casualties lost to the service by death and evacuation to the United States. The rifle caused more head wounds than any other weapon and was second only to the machinegun in relative lethal effect in head wounds. It ranked third in relative lethal effect in thoracic wounds,

TABLE 87.-Anatomic distribution (regional frequency) of wounds, by causative agents

¹Percent for dichotomy, dead versus survived, by each anatomic location and for total dead versus survived by each causative agent.

being exceeded by the machinegun and artillery shell, and second in abdominal wounds. While the rifle was second to the mortar in causing wounds of both the upper and lower extremities, it produced more fractures than any other weapon (fig. 180).

The machinegun caused fewer casualties than any other weapon, 8.4 percent. However, its percentage relative lethal effect was the highest of all weapons, 57.6 percent. It was not possible to separate the casualties produced by the 6.5 mm. weapon from those produced by the 7.7 mm. machinegun. The percentage lost to the service by death and evacuation to the United States was also the highest of any weapon, 78.1 percent. Measured by the number of patients lost to the service, machinegun wounds were the most severe among those produced by any weapon. This high degree of effectiveness of the machinegun bullet may be explained partially by close range fire in this campaign and also by the multiplicity of wounds. The percentage relative

341

FIGURE 180.-Roentgenogram of compound comminuted fracture of the humerus caused by a Japanese .25 caliber rifle bullet fired from a distance of 75 yards. This is a typical example of the explosive effect of the .25 caliber rifle bullet when it strikes bone or a solid organ. Many of these bullets caused similar damage to the arm and then passed through the chest.

TABLE 88.-Disposition of 1,337 nonfatal casualties, by causative agent

¹Defined as the beachhead perimeter on Bougainville Island.
²From hospitals on Guadalcanal, Espíritu Santo, and New Caledonia.
³Percent of total casualties inflicted by causative agent.
⁴Percent for dichotomy, first echelon, versus rear echelon of those returned to duty (= 100 percent).

342

FIGURE 181.-Roentgenogram of thoracic cavity of soldier who was prone on the ground when a mortar shell of unknown size exploded 1-yard distant. This soldier was also wounded in the arm, thigh, and both ankles. An open operation was performed, and the numerous lacerations in the lung, caused by the small fragments, were sutured and the intercostal vessels ligated. The soldier made a good recovery.

lethal effect was uniformly high for all regions of the body with the exception of the extremities.

The mortar caused more wounds than any other weapon and accounted for 38.8 percent of all battle casualties. However, its relative lethal effect was only 11.8 percent (fig. 181). The only weapon having a lower lethal effect was the grenade. Furthermore, 71.5 percent of the living wounded were returned to duty, a higher percentage than for any other weapon except the grenade. The dead and evacuated to the United States (lost to the service) totaled 28.5 percent. The highest relative lethal effect (30.8 percent) was observed in wounds of the abdomen, whereas the greatest number of deaths occurred in multiple regional involvement.

The use of artillery by the enemy in this campaign was relatively limited. Wounds caused by artillery shells, however, accounted for 10.9 percent of the casualties and were fourth in frequency. Artillery ranked fourth in cause of death (11.1 percent) and fifth in percentage lethal effect, 22.7 percent. Among casualties evacuated to the United States, artillery produced the lowest number of wounds, 13.9 percent. However, the percentage of those lost to the service by death and evacuation to the United States was 36.6 percent. While wounds

343

FIGURE 182.-Roentgenogram of thoracic cavity of soldier who was prone in a foxhole when a Japanese hand grenade exploded at a distance of not more than 1 foot from the chest wall. This X-ray shows the characteristic small fragments of the hand grenade. Most of the fragments were stopped by the chest wall, but some of them penetrated the pleura. The fragmentation of the Japanese hand grenade is irregular but usually very small.

of the extremities were frequent, only one death occurred. This death was produced by a lower extremity wound. Lethal wounds in order of frequency by regions were the thorax, head, multiple, abdomen, and lower extremity.

The grenade ranked third in wound production and accounted for 12.5 percent of all battle casualties (table 71). However, its relative lethal effect was the lowest of all weapons, 6.2 percent. Furthermore, the majority of the wounds were of a minor nature (fig. 182). The grenade was first among all weapons as gaged by the percentage of wounded returned to duty, 77.2 percent, and three-fourths of these patients were returned to duty from the first echelon. The grenade was responsible for the lowest number of casualties (22.8 percent) among those who were lost to the service by death and evacuation to the United States. Of all wounds produced by the grenade, 68.7 percent were classified as extremity wounds and multiple wounds.

Weapon Evaluation by Multiplicity of Wounds

The question has been frequently asked: Do missiles causing multiple wounds result in more serious casualties because of the number of wounds per se? The data available do not answer this question satisfactorily. Multiple

344

wounds were analyzed according to the number of different anatomic regions involved rather than by the total number of wounds. Thus, a patient with 10 wounds of the leg and 5 of the hand was classified under multiple wounds in two anatomic regions; that is, as an upper and a lower extremity casualty without regard to the number of lesions present.

Table 89 relates the casualties with multiple wounds to the number of anatomic regions involved and the severity of the wounds. The disposition of the patient was used to determine the severity of the wounds. The number of the multiple wounded casualties discharged in each echelon is tabulated by weapon. The corresponding number of anatomic regions hit is also recorded by weapon. Thus, there were 53 patients, with mortar wounds in 117 different anatomic regions, returned to duty in the first echelon. Therefore, among the patients returned to duty in this echelon, there were mortar wounds in 2.21 of the various anatomic regions per patient (table 90). The ratio of anatomic regions wounded per patient is slightly higher for each weapon among the casualties evacuated to the United States. However, the difference is so slight as to suggest that multiplicity of wounds alone is not a factor of great importance. The relatively low mortality of 3.3 percent for all patients with multiple wounds seen alive suggests that the multiple wounds per se add little to the risk. It is likely that the actual severity of the wound is the more important factor in determining death and disability. It would be desirable, however, to have data which include a count of the actual number of wounds by anatomic region in both the living and the dead.

TABLE 89.-Disposition of patients with multiple wounds as related to number of anatomic regions hit and to severity of wounds, by causative agent

¹Defined as the beachhead perimeter on Bougainville Island.
²From hospitals on Guadalcanal, Espíritu Santo, and New Caledonia.

345

TABLE 90.-Ratio of number of anatomic regions hit per patient evacuated in each echelon, by causative agent

¹Defined as the beachhead on Bougainville Island.
²From hospitals on Guadalcanal, Espíritu Santo, and New Caledonia.

Relative Lethal Effect of U.S. Weapons and Japanese Weapons

It had been the intention of the survey team to study the effect of U.S. weapons on the enemy dead. Unfortunately, this plan was found impracticable because of difficulty in obtaining the enemy dead before decomposition had occurred and also because of the paucity of team personnel. Certain local conditions prevailed which circumvented accuracy in such a study. In the first place, because of the character of the fighting and the extensive use, by Allied forces, of artillery and mortar fire, the enemy dead were frequently struck by many different missiles before the bodies could be recovered. Furthermore, it was impossible to obtain any detailed information regarding the circumstances surrounding death.

It was possible, however, to investigate the effect of U.S. weapons on a limited number of American soldiers who were wounded (table 91). There were 219 casualties (12.3 percent of the total) due to U.S. weapons in the hands of American troops. Though the Japanese used some U.S. weapons, particularly rifles and grenades, as a rule it was impossible to know when this occurred. Among Allied forces, there were 63 deaths (16.0 percent of the total dead) produced by U.S. weapons.

There were 52 casualties caused by the rifle, 16 of whom died (table 92); 19 were wounded by the accidental discharge of a rifle by a fellow soldier. Mistaken identity resulted in 13 deaths and the wounding of 6 others. Of these deaths, 8 were occasioned by the soldier seeking to relieve himself at the toilet during the night. Self-inflicted wounds, accidental or intentional, were responsible for 10 casualties, 3 of whom died. Mortar and artillery fire accounted for 54 of the wounded and 22 of the dead. Among these, 13 were killed and 40 wounded by mortar and artillery "shorts." Among the 16 casualties who were

346

wounded on patrol by U.S. artillery, 8 died. The accidental tripping of landmines and boobytraps produced 14 deaths in a total of 40 wounded. Hand grenades, other than those used in boobytraps, were responsible for 8 deaths and 4 wounded. Miscellaneous weapons including bangalore torpedoes, bombs, pistols, knives, and powder explosions accounted for 38 casualties; 7 of these casualties died.

TABLE 91.-Distribution of 219 U.S. casualties produced by U.S. weapons, by category

¹Defined as the beachhead perimeter on Bougainville Island.
²From hospitals on Guadalcanal, Espíritu Santo, and New Caledonia.

TABLE 92.-Relative lethal effect of U.S. weapons on 219 U.S. casualties