The writer, judging from data presented, favors the substitution of the judgment of the physician for the conclusion by the arbitrary rule in the filing of death certificates. This is in agreement with the procedure initiated in England, Wales, and Canada some yrs. ago. But the physician must be taught to understand what data he must furnish the statistical office and in what order they must be presented. Mere underlining of arbitrary terms seems inadequate. This instruction should begin in the medical school.