APPENDIX B 



CONSTRUCTION OF CLIMOGRAPHS 



IN the preparation of the climographs which illustrate this book 

 the first step was to select a series of cities or districts so located 

 as to cover all parts of the area to be investigated. The number of 

 such cities was fourteen in France, fourteen in Italy, fifty-two in the 

 United States, and also Tokio in Japan. Then the deaths at each 

 place during the years 1899 to 1915, or for as much of that period as 

 possible, were tabulated by months for each year. Allowance was 

 made for the fact that some months are shorter than others. This 

 was done by multiplying the figures for short months by ^^^s; ^%9^ 

 or ^Yso as the case might be. Thus figures were obtained representing 

 the actual monthly degree of mortality. 



The second step was to find how much this actual mortality differs 

 from the normal mortality for the year in question. The normal may 

 increase because of the natural growth in population. It may also 

 decrease because of improved medical and sanitary efficiency. The 

 method of obtaining the normal is illustrated in Figures 28 and 29, 

 where Chicago serves as an example. In both figures the broken 

 dotted line shows the actual number of deaths from year to year, or 

 from month to month. The straight dotted lines show the normal 

 number, that is, the number that would have occurred had there been 

 no irregularities due to epidemics, weather, and so forth. In most 

 of the cases employed in the preparation of climographs in this book 

 the normal is represented by a straight line. The exact direction of 

 the line is found by dividing the years into an earlier and later half 

 and getting the average for each half. The two averages are plotted 

 in their proper places and then connected by a straight line. In some 

 cases, however, the normal forms a curve instead of a straight line. 



