THE YEARS 1949 TO 1962 CONSTITUTED A STABLE PERIOD MARKED BY THE UNIFICATION OF POLICE FUNCTIONS. THE MODEL FOR COLLECTING STATISTICS IN THAT ERA INCLUDED STATISTICS ON ALL POLICE ACTIVITIES AND ON CRIMES INVESTIGATED THAT INDICATED THE SERIOUSNESS OF CRIMES, WHICH WERE ARRANGED HIERARCHICALLY; ECONOMIC AND NATIONAL SECURITY OFFENSES WERE ALSO LISTED. THE NUMBER OF CRIME DIVISIONS, WHICH REFLECT THE VALUES OF THE TIMES, WERE CRIMES AGAINST PERSONS (4), CRIMES AGAINST CHILDREN (5), MORALS OFFENSES (6), THEFTS (5), CURRENCY CRIMES (6), FRAUD (3), AND MISCELLANEOUS, (15). THE STATISTICAL FORMAT OF 1949 TO 1962 WAS CHARACTERIZED BY RESISTANCE TO CHANGE, CENTRALIZATION, THE APPEARANCE OF AN INDEX, THE PREEMINENCE OF SECURITY FUNCTIONS, AND CONFUSION OF THE SENSORY AND THE INTELLECTUAL CAUSED BY FASCINATION WITH FIGURES, AS WELL AS BY EPISTEMOLOGICAL COHERENCE AND INTERDISCIPLINARY UNIFICATION. AT THE SAME TIME, THE FIGURES BECAME ARTIFICIAL AND MECHANICAL, REPRESENTING A POLICE SYSTEM WHICH SERVED AS AN INDEX FOR THE PRESSURE OF CRIME ON THE SYSTEMS OF CONTROL. HISTORICALLY, CRIME STATISTICS SHOW CRIME FLUCTUATIONS FOR CERTAIN CRIMES, AS THE EXAMPLE OF AUTO THEFTS ILLUSTRATES. DECENTRALIZED STUDIES OF SEPARATE CRIMES MAY PERMIT DOCUMENTATION OF GENERATIONAL CHANGES IN VALUES, IF HISTORICAL FACTORS ARE ALSO CONSIDERED. A REFORM OF 1963 SOUGHT TO INTRODUCE A TECHNIQUE OF DECENTRALIZATIONS; SPECIFICS OF DATA COLLECTION, MODEL CHANGES, AND ADDITIONS TO OR RECLASSIFICATION OF CRIME HEADINGS WERE CATALOGUED. IN GENERAL, OFFENSES INVESTIGATED WERE SEPARATED, AND DIMENSIONAL ANALYSIS OF LARGE CRIMINALISTIC AREAS WERE SYSTEMATIZED TO IMPROVE OBSERVATIONS ON THE EXTERIOR MANIFESTATIONS OF CRIME AND TO FACILITATE SERIOUS STUDY OF REAL DISSUASIVE EFFECTS. THE EXAMPLE OF VOLUNTARY HOMICIDE STATISTICS CAN BE USED TO ANALYZE CRIME TRENDS. NOTES ARE SUPPLIED. --IN FRENCH. (KMD)