is a reflection of the seriousness of the crimes involved, but other factors may also explain this difference. However, the most notable differences concern the number of placements in police custody in each gendarmerie, as compared with placement decisions in police stations. In the latter, police custody needs to be considered on the basis of the number of placements per day. In gendarmeries, it has to be counted on a monthly basis. The cases noted in the four gendarmeries analysed ranged between 1.1 and 4.4 placements in police custody per month. Admittedly, there are some gendarmeries in which the numbers of cases of police custody are considerably higher, in particular in the peri-urban zones of large urban centres. However, police custody is, generally speaking, a relatively rare event for gendarmes and a much more routine occurrence for the police. It is easier, for the former, to adapt the service to the character of the person placed in police custody rather than the inverse, as is done in police stations: taking the person into the yard for a smoke, allowing them to eat their meals outside of the secure room etc. It is striking, for example, how rare it is to find gendarmeries that possess means of restraint for agitated persons (hitting themselves against the walls etc.); most police stations, on the contrary, are equipped with crash helmets, and even means of physical restraint, for this purpose. In gendarmeries issues of this kind are resolved through dialogue, or otherwise by means of medicalization. These small numbers explain why gendarmeries do not experience congestion of their secure rooms, allocated without distinction to offenders and persons in a state of drunkenness, without any premises reserved for minors or women, for example. There is never more than one person per “room”, which means that there is no shortage of equipment (a bed board and a blanket in each secure room): for example there are no cases of persons sleeping on the ground due to lack of space in shared cells. And in case of exceptionally large numbers, any extraneous persons are sent to neighbouring gendarmeries. The low rate of frequency also probably explains why the registers are in general much more efficiently kept by the gendarmes than by police officers, who are moreover divided into several different departments. These low numbers have their own consequences: inspectors have found “tubs” of food whose use-by date had expired. Although it is true that similar discoveries have been made in police stations51. The cleaning of blankets – in spite of much lower costs – does not appear to be much better organised in the gendarmeries. In one of those examined here, no provision was made for cleaning. The two purposes (sobering up and police custody) for which secure rooms are used explains why each of them contains a toilet, which avoids difficulties concerning access to bathroom facilities. Cleanliness is usual and bears no comparison with the generally dirty state of most police cells: “secure rooms” are much more infrequently used; the soldiers keep closer watch; persons in police custody in gendarmeries are more often allowed out of their cell (see below); finally, even when cleaning is entrusted to a private company, cleaning of the secure rooms is incumbent upon the gendarmes (the occupants are apparently sometimes asked to put the premises back into proper order before their departure). This cleaning is regular although it is regrettable that, in two cases out of four, it is carried out according to a schedule, and not after each occupation. 51 Annual Report of the CGLPL for 2011, p. 58. 48