The Evolution of the IJA Shelter Half (1899-1945)
Article about: The Evolution of the IJA Shelter Half (1899-1945) 1906: Canada Honors the Father of the IJA Shelter Half In Canada, my third home country, in the Okanagan Region of British Columbia, there i
The Evolution of the IJA Shelter Half (1899-1945)
The Evolution of the IJA Shelter Half (1899-1945)
1906: Canada Honors the Father of the IJA Shelter Half
In Canada, my third home country, in the Okanagan Region of British Columbia, there is a scenic lakeside neighborhood called “Oyama” within the district municipality of Lake Country. The name sounds quite Japanese and for many in the BC Province, familiar with the history of settlers from Japan, a Japanese name here and there won’t come as any surprise. But this Oyama was no Japanese-Canadian lumber-jack-come-mayor, but a Japanese Army Field Marshal. The area was actually named in honor of Iwao Oyama , who served as Commander-in-Chief of the Japanese Armies in Manchuria during the Russo-Japanese War of 1904-1905.
1899: Oyama’s Tent Development Mandate
The same Oyama, when he was Chief of General Staff, wrote a letter to the Minister of the Army, Taro Katsura on 20th October 1899 in which he outlined the urgent need for the Japanese army to seriously develop appropriate gear and know-how for camping in the field. This letter served to launch the IJA’s development work on tents and shelter halves, so I will translate the whole letter below.
“The current style of warfare that requires us to amass large armies in one area for deployment would inevitably lead to increased needs of setting ourselves up in the field, even when operating in well populated areas. This becomes even more obvious when we imagine the vast and sparsely populated lands in which we will need to operate in the future. In such cases, not only the army itself, but even warehousing and hospitals in rear depot areas cannot rely any more on locally available structures. In addition, the area is bitterly cold during winter and sweltering hot in summer, which will in no time take its toll upon our army. For this reason it is of paramount importance that we thoroughly prepare already in peace time, the necessary hardware to sustain ourselves in the field.
We cannot afford to overlook the fact that the Russians have already been making great effort in this regard, conducting frequent field camping exercises around their garrison areas to improve their gear and harden their troops to being at home even when setting up in the fields.
For these reasons I would like to urgently discuss how we may go about conducting a study on the various tents used by the armies of other nations and determine what is best suited to accommodate our armies in the field. “
1895: The Monday Meetings
Oyama’s fear that the Japanese Army was inadequately prepared for living off the land, away from home was a widely recognized sentiment since experiencing the Sino-Japanese War of 1894/95. Up until then the concept of war in Japan was limited pretty much to civil war, and soldiers were not used to the idea of “roughing it in the bush” in the middle of nowhere. So after the war had ended in November of 1895, Oyama formed a “Special Examiner’s Committee for Improving Uniform, Field Accommodation, Equipment and Portable Rations” （被服装具陣具及携帯食料改良審査委員会）when he was Minister of the Army. The committee was headed by a general, reporting directly to the minister and committee members were a total of 26 officers from the various branches of the army, including medical officers. They met up each Monday to discuss matters like canteens, mess kits, field rations, field kitchens, back packs, field medical kits, footwear and it was also this committee that took up the development of tents and shelter halves. They not only discussed, but prototyped and tested equipment. This committee was active for roughly 20 years until all such R&D functions were centralized. Back then, artillery and field engineers also had to have their own meetings to develop weapons until they joined up in 1903 under the Army Technical Assessment Department (技術審査部), which further evolved into Army Technical Head Quarters in 1919.
Half, Quarter or One-third of a Shelter?
In Japanese it is called Keitai Tenmaku (携帯天幕), which means portable tent. In modern day Japanese a cell phone is called a “Keitai”, which is short for “Keitai Denwa” or portable phone. Tenmaku consists further of the word “Ten” meaning “Heavens” and “Maku” meaning “Curtain/Screen”. So in Japanese it was literally a “portable overhead tarp” without any fractional concept like half or quarter. Actually the army regarded 3 as a full set for sending samples or calculating material usage, so it could even be a one-third shelter, but as I have to call it something in English, I will call it a “shelter half” as per the 1944 US Technical Manual on the Japanese Military Forces. That manual even says “a standard method of pitching requires 28 shelter halves”. So a one-twenty-eighth shelter even? Sometimes Japanese makes more sense than English.
March 1903: the One Man Tent
3 years after Oyama had ordered development of individual field shelter, the Chief of the examiner’s committee finally reported to the Minister of the Army on 15th August 1902 that they had surveyed what other nations used as tents, selected a design and had the 1st and 7th Infantry Divisions do some field tests and finally settled on the following specs, which was officially introduced 7 months later on 19th March 1903 as Army Ordinance 13.
1. The tents for the troops were self-portable types to be carried by individuals, NCOs and below.
2. Each individual was to carry the following.
a. Shelter half 956 g
b. 2 Tent pole sections 330 g
c. 2 Tent pegs 94 g
d. 1 length of Rope 15 g
Total 1,395 g
3. The shelter half was square shaped with each side measuring 1.5 m. In order to facilitate joining of more than 2 sections, 2 sides had 7 looped cords, and the other 2 sides had aluminum grommet reinforced holes to take the cords. Joining 2 panels together was easily achieved by weaving the 7 looped cords one by one through the 7 grommets on the adjoining panel and anchoring down the last loop with a tent peg (See diagram below). In the 4 corners and center of each side were anchoring cords. There were grommets in the 4 corners for the tip of the pole. Other armies at the time either used square or diamond shaped (Austria) shelter halves, but the majority were square, as the shape lent itself easily to various connecting combinations when erecting tents. Particular attention was given to the German and Russian designs in finalizing the Japanese design. The improvement examiner’s committee was astute enough to get the Russian training manual on pitching tents fully translated for study, so the IJA was well prepared for the coming war with Russia in 1904/05.
4. The fabric was a hemp/cotton mixed weave dyed in a light earth brown. The fabric tested included hemp, cotton and hemp/cotton mixed weave, and for each they assessed weight, shrinkage, quickness of drying, tear resistance in longitudinal as well as lateral directions. The cloth adopted was made by Hokkaido Seima (hemp manufacturing) Company (北海道製麻会社).
5. The pole was in Japanese oak or similar wood and collapsible into 3 sections of 45 cm in length, fitted with steel hardware at the connecting points (Hardware was initially planned to be all aluminum in order to save weight, but as the private sector had not yet developed much for this metal, all aluminum parts would have had to be manufactured by the Tokyo Artillery Arsenal. However, aluminum hardware for the poles had to be switched to steel because of poor strength)
6. The pegs were of Japanese oak or similar wood with a length of 27 cm and with one curved side to offer better bite. Robust material needed to be chosen as wood for the pole sections and pegs, but there they encountered difficulty, as most types could not take the stress well. However, degreased Japanese Oak was light and strongest among those tested.
7. The rope was a hemp rope of 2 meters.
Carry method (fig 4)
Those with backpacks were to role up the 2 poles and 2 pegs in the shelter half into a bundle, the length of the pole section and fix it to the upper end of the back pack to the outside of the overcoat as shown in fig 4. Mounted individuals were to fix this bundle on top of their coat roll at the rear of the saddle.
Wear as rain cape (fig 5)
When wearing the shelter half as a rain cape the section of cord “a-c” was to be tied around the neck and the cord “b-f” around the waist. Corners g and h were to be brought to the front and by tying the corner cords to the side cords, were to be fashioned like a sleeve.
Pitching tents of various sizes
For all tent structures, the bottom end of the tent pole needed to be anchored in the ground. Rocks or wood blocks were to be used as a base when erecting the pole.
1. Pitching a 1 person tent. 2 diagonal corners were to be pegged down and one corner raised and tied to a tree, etc as in fig 6
2. A 2 person tent may be erected as in fig 7, but in this case the poles required improvisation.
3. A 3 person tent was pitched as in fig 8. In this case, the angled section A was to face upwind.
4. A 4 person tent was as in fig 9, and was particularly suitable for hot summers. When having 6 people together, a spacious summer tent as shown in fig 10 could be formed.
5. A 6 person tent could be pitched as in fig 11. In this case one end was to have steps for getting in and out, and if the other end could be closed off against a bush, etc, the structure could easily accommodate 7, in which case the 7th shelter half could serve as the entrance cover.
6. A 24 person tent with a central hearth could be formed as shown in fig 12. First, the hearth was to be surrounded with 8 shelter halves, and the 2 inner corners of each square were to be connected and held up with the poles. Then an octagon or circle was to be drawn around the outer circumference of the 8 squares which should be built up into a circular mound of approx. 1.5 m height with a 75 cm wide opening at an appropriate location. When the surrounding mound was finished, the triangular gaps between the initially placed 8 squares were to be closed by using another 8 shelter halves, which were to have the center grommet of the inner side pierced by the 8 poles in the center and the outer corners anchored into the surrounding mound. The 2 free inner corners were to be tucked in. A further 8 shelter halves could be used inside as required for partitions, etc.
7. A 38 person cold climate shelter could be formed as per fig 13. This structure required 24 shelter halves and 8 poles. Firstly, any snow was to be cleared off the surface. Then 8 squares were to be connected into 2 rows of 4. Another set with 8 squares in 2 rows of 4 was to be prepared. These two large 2x4 combinations were joined as one big piece, but the 4 squares in the middle were to have the inner corner folded over as smoke vent. The poles were to be placed at positions A, B, C and D. The poles were to be supported in position by ropes attached to the tips and the other end anchored with pegs. The other 4 poles were to be used where needed to keep the structure from sagging. Then 12 squares were to be hung around the 4 sides with each to having the lower edges pegged. After they were anchored, some snow was to be thrown on along the rim and stamped down with feet. At the 4 corners shelter halves folded into a triangle were to be hung, the one downwind to be kept as a entrance flap.
8. A 30 to 40 person tent surrounded by a circular mound is shown as fig 14, which required 15 shelter halves. A pole in the center was stood where the hearth would be and a circle was drawn around it at a radius of 3 paces. Along this circle, 8 squares were to be placed side by side with the inner corners joined and supported by poles. At a downwind position an entrance was to be kept open at the mound. The mound was to have a depth of 75 cm and height of 1.5 m. Once the mound was complete, the outer corners were to be pegged down into it. Folded halves were to be used to close the triangular gaps between the 8 squares. In order to prevent smoke from filling the tent, 8 squares were to be hung down in circular fashion along the inner rim. These could be lifted up when there was no wind. Also a shelter half was to be hung at the entrance to serve as a flap. Altogether, this structure required 24 shelter halves. The remaining pieces were to be used inside as flooring. This structure could accommodate 30 to 40 persons lying down.
This 1903 Model did not get any Type designation, so for this story, let’s call it the Model 03 hereafter. All 1903 Model shelter halves were produced at the Tokyo Artillery Arsenal at least till end of 1912.
Last edited by nick komiya; 11-14-2015 at 04:47 PM.
1904: The Russo-Japanese War and the Shelter Half Crisis
One year after the launch of the Model 03 shelter half, the Russo-Japanese War broke out (8th Feb. 1904-5th Sept. 1905). They did manage to send their troops to the continent equipped with the new shelter halves, but by July of 1904 an appalling number of replacement orders were coming in from the field for the new model mess kits as well as the shelter halves. The army was completely swamped with orders for resupply, but could barely supply the new troops shipping out and had absolutely no room to deal with replacement demand for field damages. There was already a conservation policy in effect, prohibiting peace time use of mess kits for cooking, but even though there was now a war going on, troops had to be told to avoid cooking with mess kits as much as possible and not to count on any replacements if they could not make their gear last. For the shelter halves, one problem was that they faded and lost water repellency. Finally the army had to send workmen to the field army in Manchuria to re-dye and renew the water repellent finish on 50 thousand shelter quarters. In August of 1904, 22 workmen left from Tokyo with a daily quota of 2000 shelter halves to refinish.
There were also problems of tent poles and pegs breaking. To cope with these issues they decided to incorporate improvement suggestions and revised specs on 1st June 1904 through Army Ordinance 101. Changes made were as follows
1.The small wood screws used to fix the steel hardware on both ends of the pole were known to fall out and were replaced with rivets
2.The wooden poles were too shallowly inserted into the metal cups. For proper strength, the depth of this coupling would need to be at least 1.5 times the diameter of the pole. So the 16 mm joint was increased to 30 mm by stretching the steel cap to be deeper. At the same time the metal was made thinner not to increase weight, due to this change.
3.The steel pole ends were galvanized to prevent rusting.
4.The poles with different hardware on the ends should nonetheless have the wooden pole inserted to the same depths into the hardware to ensure equal joint strength.
5.The pegs were initially designed to have a thickness of 10 mm at the head and taper down to 7 mm at the tip. However, because the pointed end of the pegs tended to become thinner for wear anyway, the whole peg was made to be 10 mm thick, which also had the benefit of adding overall strength and making them easier to produce.
6.The center seam of the shelter half was changed to reverse stitching in order to increase durability.
7.Method of hemming the edges was simplified in case of selvedge use to reduce material waste and yet keep the same finished size despite variances in the size of the cloth panel.
8.The cord for tying at the waist got simplified.
In mid April of 1905 another dispatch of workmen to the continent became necessary to re-dye and renew the water repellent finish again. This time they would need to serve 5 armies and refinish a total of 280 thousand shelter halves at a pace of 2000 per day again. As can be seen, this alone made the item extremely high maintenance and was a bitter and expensive lesson for those who developed the gear. The water-proofing process is described in detail in a 1917 uniform care manual, and involved saturating the fabric in a solution of Marseille soap and then in an alumina solution, which allows the alumina to react with the soap in the fabric to form a water repellant substance. This same process was used on rain coats and bread bags.
Another crisis that the shelter halves had to overcome during the Russo-Japanese War was the manufacturer of the cloth in the Northern Island of Hokkaido completely running out of the hemp/cotton mix, so it was necessary to substitute it with a heavy cotton duck. A memo dated 28th March 1906 stated that the Tokyo Arsenal had stocked up on this cotton duck during the war, and though now that the war had ended and production should return to the specified hemp/cotton material, the arsenal would first need to use up 10 thousand shelter halves worth of heavy duck to prevent all that stock from becoming dead stock. The memo further stated that water repellency and robustness of the material proved in no way inferior.
1910-1912: Tweaking the Fabric composition
The Model 1903 fabric being a mixed weave meant that the lateral threads were cotton and longitudinal threads in hemp, which made the weaving a difficult process and because the threads were thick, the end product tended to be heavy. They now attempted to kill several birds with one stone by reversing the thread choices for lateral and longitudinal to facilitate ease of production and by using thinner threads reduce weight and increase weave density at the same time, which helps to increase water repellency. In combination with this new material they wanted to try out the new indanthrene vat dye known for its fastness to light and washing. This dye was a new invention, first patented in 1901. The weight reduction from this change was 124 g. In order also to test the fastness of the indanthrene dye, they made three types of shelter halves. Two types using the new weave pattern, one dyed the conventional way (red iron oxide dye) and another using indanthrene. The third was the current spec shelter half as a bench mark.
This field test ran approximately 1 year from early April of 1910 to end of June 1911 and was conducted by the 2nd and 3rd Imperial Guard Regiments. The outcome was that the new weave had successfully reduced the overall weight, improved water repellency and even improved fastness of the conventional dye. The new weave using indanthrene looked sharp at first and showed good water repellency, but by the end of the test, water repellency was drastically reduced and the color was also more faded than the new fabric dyed in the conventional manner. The new weave had successfully maintained water repellency throughout the test and some examples even showed no diminishing of this attribute at all till the end. Indanthrene actually had an adverse effect on water repellency and also faded quicker on the new cloth.
On 20th September 1912, the specification of the fabric was officially changed to make hemp the lateral element instead of the longitudinal.
Aftermath of WW1
WW1 was a fairly eventless war for Japan, as the 4700 German soldiers in China had surrendered all within 1914, but it did not escape notice of the army that a new age of air wars had arrived. This resulted in the IJA conducting some tests on concealment of tents from enemy aircraft. In 1916, color comparison tests were carried out on conventional shelter half cloth against 2 types of cloth with green and brown woven stripe patterns. They looked down at them from observation balloons to see whether there was any improvement in the camouflage effect. However, they must have not noticed any meaningful improvement, as no spec changes came out of this attempt.
June 1937: Quality Problems and the Finger Tip Tear Test
A memo from the commander of the 16th Infantry Division addressed to the Minister of the Army reported that many shelter halves in their inventory tested, by the method as advised by the central clothing depot in 1934, had proved to be substandard quality and needed to be disposed. The test was simple, whether one could easily rip the material by holding it between finger tips. A list attached to the report shows material defects by production year. Interestingly the majority of their inventory came from 1907/08, but some examples went back as far as 1904. Either way, all examples that failed were of the old pre-1912 fabric, but it is interesting that they had kept inventory as old as 33 years. The total number they had in stock at the division came to 49,848 shelter halves of which 20,909 or 41% had failed in the finger tip rip test.
May 1939 – February 1940: Testing of Triangular Shelter Quarters
In the meanwhile, in Europe, the Germans had already concluded their prototyping studies of a triangular shelter quarter in May 1926, which got introduced on 7th March 1927. This further got its splinter pattern camouflage print added from 26th June 1930 to become the “Zeltbahn 31”. The Germans and the Japanese provided each other with samples of army uniforms and equipment on a fairly regular basis, so the triangle idea was also assessed by the IJA right before WW2.
The Japanese Army had added their twist to the triangle design, so in addition to being a tent and rain cape, it was to be used as a backpack. The poles, pegs and rope to be used for these prototypes were the same as those of the M1912.
By mid May of 1939 the cavalry school, field artillery school and combat engineer schools received 100 pieces each of the triangular models. The infantry school got 150 pieces and the Imperial Guards Division received 250 for testing in comparison to the M1912 model. Testing was to run until end of February of 1940 with a final report due in March 1940.
I will omit detailed descriptions, as readers are likely to be familiar with the German M31 model. However, it appears that the joining of panels was not done by buttons through button holes like the German model, but by press studs with the button side made of wood. Also unlike the M31 it had a flap at the neck hole in center which served as a collar when wearing as a rain cape and as a lid flap when using as a back pack.
I could not find the test results, but history proves that this model was not adopted in the end, though the army did get registered rights for the design on 10th September 1940.
1941 Leather reinforced holes instead of aluminum grommets
The last recorded change to shelter half specs was announced on 20th November 1941 as Army Ordinance 82, which added that leather reinforcements would also be accepted in lieu of aluminum grommets. This happened to be the same ordinance that also introduced the tie cords for the canteen stoppers instead of the leather strap and buckle system. In the 1943 army spec book this 1941 version is referred to as Model Ro (Model B). In this case, the M1912 with aluminum grommets would have been Model I (Model A). Collectors seem to associate leather tabs on shelter halves to be a feature adopted only after 1944, but they were already allowed on an optional basis from late 1941.
Also within the same spec book is a cotton version of Model A, in case hemp ran out as it did during the Russo-Japanese War. This was set up as a backup spec on 28th April 1938.
On the surface, the IJA shelter half hardly seemed to change over 40 years of its existence, so books discussing Japanese field gear devote only a paragraph at best to the subject. However, even an item that appears to be a total bore had seen a fair amount of drama in its history. The major milestones, looking back, are as below.
19th March 1903 Launch of the Model 1903
1st June 1904 Last change in design (cord, seam, pole sections and peg thickness)
20th Sept. 1912 Weave pattern of hemp and cotton mix reversed
28th April 1938 Cotton version set up as back up spec
20th Nov. 1941 Model B shelter half added as production variant depending on material availability
Great article Nick! Puts to rest some misconceptions I had and fills in many blanks. It's a really interesting piece of military field equipment.
Terrific that you are a fellow Canuck couldn't be happier to hear it, even if we have to share you.
In Field equipment, kit and other
In Field equipment, kit and other
In Field equipment, kit and other
In Field equipment, kit and other
In Field equipment, kit and other