The public remained unconvinced, as well they might with approximately 300,000 dead unaccounted for. On 17 March 1922 a Mr Chapman, residing in Mailly Mallet wrote to James Gillies, Minister of Lesmahagon, whose son was missing at Serre: "Today a Belgian found a body and reported, on his being questioned he admitted that an old cigarette case was there but had been thrown away. Well, I threatened and frightened him and at midday it is produced and inside it is a photo of the poor fellow with his name and address." On 26 May 1922 he wrote again of "Hundreds if not thousands" of British bodies being exposed. He deemed the two franc reward inadequate when the Frenchman might lose seven or eight francs walking seven kilometres to inform someone of the discovery of remains. He wrote of "The searching and desecration of the dagos who are doing the work of clearing the ground... Last week I pointed out some bodies with the result that an action was immediately taken, identifications secured at least two cases... The whole circumstances are a disgrace to our nationhood." Ghillies then wrote to Captain Elliot MP: "I visited the Somme district in November last. The ground for the most part lay as at the Armistice - & thousands of unknown British soldiers are being brought up as the work of restoration proceeds."70
It is an inescapable conclusion that definition of the army's task as ended in 1921 was arbitrary. Worthington-Evans's statement gave the impression that some turning point had been reached. It had not. The withdrawal of the army was more a response to diminishing manpower, problems on the ground, and finance. The task was simply passed to the IWGC. In April 1922 E.A.S. Gell reported nine gangs of 30 men and 20 gangs of 10 men working at Ypres; 150 men "(mostly Poles) working under an intelligent French foreman,"74 at Neuve Chapelle, and 25 gangs working south of the Vermelles-Hulloch Road and on Hill 70 at Loos, this being merely 3 of the 8 areas still being searched. As time went on, the IWGC relied more and more on local reporting as ground was levelled, drainage dug, and roads created. Between 1932 and 1936 4,079 bodies were recovered (7% in Belgium, 93% in France). Fifty-two per cent had been found by metal searchers; 30% by farmers/others; 18% by French government search parties.75 The figures for body recovery alone indicate the incompleteness of the task in 1921. 28,036 bodies were found between 1921 and 1928 (with 25% identification),76 and approximately a further 10,000 up to 1937.
The sheer physical difficulties of the work were significant. On the devastated battlefields much initial effort had to be directed into erecting accommodation and providing supplies. The weather added to difficulties - on 20 January 1919 frost stopped work. Up to this point five to six men were required each working day to exhume each body, transport it to the cemetery and re-inter it. When work resumed on 17 February 1919 nine men were required per exhumation per day. By 14 March 1919 there had been only 1,750 exhumations (excluding Canadian efforts). Manpower rapidly became a problem - with demobilisation, volunteers began to disappear. An undated memo of a meeting at which both Winston Churchill and Field-Marshal Haig were present, noted the 33,000 "labour men surplus at home who are retainable," but records the decision to pursue the route of volunteers.37 It was estimated in March 1919 that 12,000 men would be necessary. This was increased the following month to 15,000 Labour Company personnel, 1,500 Cemetery Party personnel and 1,787 DGR&E personnel. By 17 May Major-General Burnett Stuart was requesting 15 more "grave registration squads" from England.38
Furthermore, for a certain broad class of acyclic posets S, we develop inverse probability transforms from [0,1) into S and show how to utilize them to explicitly construct systems (Xa: a in A) with the monotonicity property from a single uniform random variable U.
We develop an inverse probability transformfor a certain broad class of posets S,and use it to explicitly constructa system (Xa: a in A) realizing the monotonicityof a stochastically monotone system when the two notions of monotonicityare equivalent.