Why Did Mine Dogs Work for the Brits but Not for the Yanks in World War II?

The British deployed over 60 mine detecting dogs to the European theater, Holland in particular, and were very successful at uncovering mines along rail lines and at transport locations. The Americans used dogs in Italy, but quickly labeled the program a failure. The British primarily used a reward method of training, while the Americans primarily used a “repulsive” electric shock. Suspicions have been raised that the two approaches might explain why one country succeeded and the other failed, but there were other differences between the programs.

A cartoon drawn by a platoon commander in the English program for the lead trainer may provide details about the English program for training the dogs, including the possibility that a sort of scent lineup was employed.

U.S. Mine Dog Program in World War II

A study by Anna M. Waller completed for the Department of the Army, Office of the Quartermaster General, describes the effort to develop mine detection dogs in World War II. The study, written in 1958, says that a number of stateside locations were used to train war dogs, including the Front Royal Virginia Quartermaster Remount Depot, Fort Robinson Nebraska, Camp Rimini Montana, and the San Carlos War Dog Reception and Training Center California. Mine detection dogs were trained at the San Carlos, California, facility. The other types of specialized dogs were sentry, sled and pack, messenger, and scout dogs.

By far the largest number of dogs were trained for sentry work, a total of 9,295 for both the Army and Coast Guard. For scout work, 571 were trained, all for the Army, which also had 263 sled and pack dogs and 151 messenger dogs. The smallest number, 140, were trained for mine detection. The organization of war dogs did not happen until near the end of the war. Waller states:

“It was not until March 1944 that the War Department authorized the establishment of Quartermaster war dog platoons and issued special Tables of Organization and Equipment (T/O & E) for that purpose. Originally a platoon consisted of twelve scout dogs twelve messenger dogs, one mine detection dog, one officer and twenty-six enlisted men. Three months later, however, on the basis of early theater experience, the mine detection dog was eliminated and the number of scout dogs was increased to eighteen, while the number of messenger dogs was reduced to six and the number of enlisted men to twenty. Fifteen Quartermaster war dog platoons were activated and trained in 1944, and all were shipped overseas. Seven of them saw service in Europe and eight in the Pacific.”

Waller’s study attributes the North African Campaign with creating the perceived need for mine detection dogs (“M-dogs”), which was where the enemy first used non-metallic plastic and wooden mines, which metal detectors could not find. A detection unit was ordered activated in November 1943. Dogs were taught to detect buried objects so that they could detect both metallic and non-metallic mines, anti-tank and anti-tank personnel mines, trip wires, and booby traps. Dogs were “taught to indicate the position of a buried mine by sitting down from one to four paces from the concealed objects.” If a dog detected a trip wire or booby trap, the dog was trained to halt or refuse to advance. Waller says the following of the training:

“A light electric charge was concealed in the trap and the dog was shocked when he came in contact with it. This was done to teach him that there were objects in the ground which would hurt him. When he had learned this, his fear of being injured made it possible to teach him to shun objects foreign to the terrain and to rely on all his senses in trying to detect them.”

The program was not successful. Waller summarizes what happened:

“The enthusiasm with which this training began later turned to disappointment. Only two war dog mine detection units were activated and trained. Both were sent to North Africa, where the animals failed to prove their proficiency in locating mines when used on typical German mine fields. The dogs had been tested in the United States and pronounced excellent detectors but when tried out in North Africa under battlefield conditions they fell far short of attaining the standard of efficiency that had been established by the Corps of Engineers. In two tests in September 1944 the dogs located only 51 and 48% respectively of the mines planted. Inasmuch as the discovery of at least 90% was considered essential to make a method of mine detection practicable, it was decided not to employ the dogs. Both units were deactivated and mine-dog training was discontinued.”

Michael Lemish, in his invaluable book, War Dogs: Canines in Combat, adds considerable detail to the American use of M-dogs in World War II. He notes that two methods, an “attraction” and a “repulsion” method, were used to train the dogs, but mostly the latter. The repulsion method used an electric shock. About a hundred dogs were trained at Cat Island, south of Richmond. A demonstration at Fort Belvoir, Virginia, consisted of dogs detecting mines in an area about 12 ft. square, with mines—wooden, plastic, and metallic—planted 18 inches apart. The dogs missed 20% of the mines, and alerted at another 20% of spots where no mines were found, but the results “impressed the reviewing officers greatly.”

Lemish writes that Col. Lee A. Denson, Jr., of the War Department’s General Staff Corps authorized training of at least 500 dog teams for mine detection. The 228th Engineer Mine Detection Company was activated at the Cat Island Dog Center. A company of 120 men and 100 dogs shipped to North Africa in May 1944, then were moved to Naples, Italy.

The dogs, which had not received refresher training for a lengthy period began probing minefields and a substantial number of casualties resulted. Further training and testing produced dismal results, with dogs detecting only about 30% of planted mines. The program languished and was finally abandoned. After failing to impress General Patton’s staff, the platoons were returned to Naples, then to the U.S., where the 228th was deactivated. Lemish notes that the dogs were not trained to work under fire. In a footnote, Lemish quotes the veterinary officer for the 228th as writing in January 31, 1945, that dogs “[u]nder fire are absolutely worthless as sentries, messengers or mine detectors, because they become neurotic very quickly, and shell fire can drive them crazy with fear.”

British Mine Dog Program

The British experience with canine mine detection was considerably more positive.

H.S. Lloyd, who was head trainer at a school for war dogs during World War II, wrote about the British mine dog program in a book published in 1948. He describes the resistance of some scientists to the idea of detecting unexploded explosive scent because of the belief that there was no scent, but resistance was overcome from a pre-war situation which involved use of a Labrador to detect leaks in a BBC cable at Cooling Marshes. The underground cable had a number of leaks over four miles, and the only way to repair it seemed to be to dig up all four miles. Then it was proposed that a particularly pungent solution be pumped through the cable, which was several feet under marsh land, and a dog be taken along the path of the cable. The dog alerted to the places where the “stink” came out of the cable, and all the leaks were found and repaired. Apparently, reference to this pre-war event convinced military authorities that mine detection dogs might work.

Lloyd writes that in training, an “association of ideas” method and “attraction” rather than “repulsion” were used. In the end, “a small percentage of dogs by reward, could indicate almost faultlessly the presence of mines buried from 3-12 inches deep no matter for what period.” Speculation was that recently buried mines still had gaps in the soil through which scent could travel up, whereas older mines reacted chemically with the soil, producing a scent that ascended to the surface.

British dogs had to be “battle inoculated,” working with the sound of Bren guns (a type of machine gun), heavy explosives at close range, and swooping aircraft. Other distractions included pinioned rabbits, sheep, game, lumps of meat, even bitches in heat, some of which are indicated in the cartoon, drawn for Lloyd by a platoon commander, J.R. Davison.

Dogs worked in platoons, three sections of four dogs and handlers, with a fourth section in reserve. The picture above shows part of a mine detection platoon. Mine detection was conducted in darkness. Before entering a danger zone dogs had their harnesses adjusted, an association of ideas that meant that work had begun. Lloyd explains how it worked:

“A front the width of a tank track could be ‘swept’ by four dogs working in echelon formation for safety purposes. Twenty to thirty minutes was the maximum period man or dog could concentrate on this work. Dogs were also employed for reconaissance patrols for mine-detecting in a similar manner to that in which a blind man’s dog functions, on a lead of some 2 ft. ahead of his handler clearing a pathway rapidly to enable the patrol to follow in safety. When the dog, working carefully with his nose suspected the presence of a mine he immediately sat a few inches short of the danger spot. His handler by prodding then confirmed the indication given by the dog, marked the position of the mine clearly so that those who followed him were conversant with the situation, and then rewarded the dog with a piece of meat carried by the handler for this purpose.”

Four mine detecting dog platoons, a total of about 64 dogs, were employed by the Royal Engineers in the North Western European Theatre of Operations between July 1944 and November 1945, meaning that they continued to be used to find mines after the war was over. Cross breeds, Alsatians, Retrievers, Collies, and others were used, and some handlers had “a distinct preference for bitches.” Dogs checked railway lines, nearly 100 miles in Holland. They were used in dumps and transport parks. Thousands of mines were detected by the platoons.

Lloyd was familiar with the failure of the American experience, which he suspected to be due in part of the use of a “repulsion” method of training in which the dog received a slight electric shock when finding a mine in training, resulting in an alert to an overly broad area. That may mean that, stressed from the shock system of training, the dogs were alerting too quickly, not getting close to the target odors. Lemish also points out that the American commanders may have had unrealistic expectations about how the dog platoons would operate.

Evidence from a Cartoon

The cartoon drawn by Davison and presented to Lloyd is meant to provide some entertainment and not to be a record of training procedures used at Potters Bar. Nevertheless, Davison titled his drawing, "New Method of Testing Scent Discrimination of War Dogs." (By clicking on the picture in most browsers a larger version of the cartoon should appear, making it easier to read some of the text; a second click may make the cartoon even larger.)

The cartoon shows a field, with a farmhouse in the near distance and a small village further away. Five pipes in a row lead to holes in a high wall. At the far end of each pipe a scent is represented as being pushed through the pipe by a fan, an extension of an exhaust pipe, or in three cases, a hand-held bellows. The five scents appear to be, from left to right, (1) a dirty sock, (2) perhaps a block of cheese, (3) a bottle of gin, (4) two mines on the ground, and (5) truck exhaust. Only three openings in the wall are shown. Again from left to right, but correlating the numbers for each end, (3) a soldier dog-like on hands and knees smelling gin, (4) a talking dog alerting to the mines (speaking to a friend), and (5) a dog coughing in the heavy exhaust coming out of the shaft. One dog says "eeny meeny miny mo," perhaps a depiction of a dog that was particularly random in his alerts when faced with a choice of five odors. Meanwhile, there are distractions in the foreground, including a cat (looking out at us to announce his conclusion that the whole business is "utter nonsense"), a greyhound wearing a Greyhound Racing Association vest chasing a rabbit and being chased by a squirrel, a bone suspended from a pole, dog biscuits on the ground that a black terrier is failing to ignore, a discarded food tin, and perhaps an opening to a sewer.

Drawing conclusions from a cartoon full of “in jokes” is risky, but it seems likely that dogs were trained to distinguish the odor of mines from other odors, and to do so in the presence of a variety of distractions. The drawing suggests that a sort of lineup system might have been used, though the term does not occur in any of the sources I’ve found so far. It does appear that the British resolved many of the issues in explosives detection well ahead of the Americans.

Sources: A.M. Waller (1958). Dogs and National Defense. Department of the Army, Office of the Quartermaster General; S.L. Smith (2001). Mine Detection Dogs at Work, The Journal of ERW and Mine Action, 5(1) (the April 2001 issue of this journal contains four articles on mine detection dogs, including descriptions of their training and their use in clearance efforts in Lebanon); B. Vesey-Fitzgerald (1948). The Book of the Dog. Nicholas & Watson, London (Chapter: The Dog in War, by H.S. Lloyd, at 176); M.G. Lemish (1996). War Dogs: Canines in Combat. Brassey’s, Washington, DC.

There are discrepancies between the accounts of Waller and Lemish that I have not attempted to resolve, such as the location of training M-dogs. Most likely this comes from their having obtained different sources.

Thanks to Richard Hawkins and L.E. Papet for corrections and suggestions.