172 BUREAU OF AMERICAN ETHNOLOGY [Bull. 197 



rates shown in table 34 point clearly to a rapid decline in overall 

 mortality levels, particularly in the early years of life following in- 

 fancy and in the early adult years. So long as the Navaho continue 

 to maintam the high levels of fertilitj^ vfhich they presently manifest, 

 continuation of the present mortality declines may well produce an 

 unprecedented rate of increase in this population in the future. 



The age-specific death rates that are shown in table 35 serve to 

 indicate both the apparent recent trends in Navaho mortality and the 

 necessarily approximate nature of all such rates, owing to the small 

 size of the population at risk, and the approximate nature of the basic 

 population totals. Four sets of rates are shown in table 35. The first 

 set is derived from the average of the age-specific deaths reported 

 among the Navaho for the years 1944 through 1946. The correspond- 

 ing population bases are obtained from an estimate of the population 

 at midyear 1945. The second and third sets of rates are derived from 

 the average of the age-specific deaths reported for the years 1948 

 through 1952. The base population used in deriving the second set 

 of rates is the Navaho population as enumerated on April 1, 1950, 

 taken without adjustment. The base population used in deriving the 

 third set of rates is an estimate for midyear 1950, w^hich corresponds 

 closely to the official Bureau of Indian Affairs estimate of the Navaho 

 population at this time. These two sets of rates, utilizing the same 

 numbers of reported deaths, are presented as a further illustration 

 of the variation which can be expected in connection with prevalent 

 uncertainties regarding the size of the base population. The fourth 

 set of rates is derived from the average number of age-specific deaths 

 reported during the years 1953 through 1957, using an estimate of the 

 Navaho population at midyear 1955 as a base. 



The range of "chance variation" (using 95 percent confidence 

 limits) was also calculated for each of the rates shown, in order to 

 indicate the amount of variation that is associated with the small size 

 of the populations at risk.^^ 



The major trend to be noted in the mortality of the Navaho between 

 1945 and 1950 appears to be a marked decline, for each sex, in the 

 broad age group from 5 to 44. The death rates of this same group 

 display an accelerated decline between 1950 and 1955. Taking into 

 consideration the range of chance variation that is associated with 

 these particular rates, the decline between 1945 and 1955 is evidently 

 significant.^^ 



" The broad age groupings used in presenting these data were selected to minimize the 

 variation associated with the small size of the populations at risk, while at the same time 

 giving some indication of the trends in the reported death rates during this period, for 

 different age-sex groups. 



13 The lowest "probable" death rate for Navaho males 5 to 44 in 1945, as shown in table 

 35, Is 4.3 per 1,000. The highest "probable" death rate for the corresponding group in 

 1955 is 3 per 1,000. This decline is even more marked among Navaho females in this age 

 group, from a probable low of 4.8 per 1,000 in 1945 to a probable upper limit of only 1.9 

 per 1,000 in 1955. 



