CONTROL OF THE PUBLIC MILK SUPPLY 461 



observations, when milk had been kept cool, the bacteria did not appear to injure 

 the children over 3 years of age at any season of the year, unless in very great 

 excess. 



"7. While it is true that even in tenements the results with the best bottle 

 feeding are nearly as good as average breast feeding, it is also true that most of 

 the bottle feeding is at present very badly done; so that, as a rule, the immense 

 superiority of breast feeding obtains." 



As a whole these extensive studies showed a less definite relationship 

 between dirty milk and infant mortality than previously had been be- 

 lieved to exist and they encouraged many other investigations of this 

 question in this country, a few of which may be noted. 



It has long been claimed that the infant mortality was lower among 

 breast-fed than among bottle-fed babies. The statement has been made 

 that 80 to 85 per cent, of all infant deaths occur among bottle-fed babies 

 but even if this is true it does not necessarily indicate that the percentage 

 of deaths among bottle-fed infants is excessive for if it should be true 

 also, that 80 per cent, of all infants are bottle-fed, this percentage of 

 infants' deaths is what would be expected and so is not an arraignment of 

 bottle feeding. Davis has reported the results of a study by the Boston 

 City Board of Health to determine the percentage of babies that are 

 bottle-fed and has made some deductions therefrom as to the influence 

 of bottle feeding on infant mortality. The work is based on the replies 

 to a circular letter of inquiry and on supplementary information obtained 

 by board of health nurses. In all there were 736 replies. Table 118 

 shows the nativity of the mother and the number of living babies at 

 various age periods from 2 weeks to 1 year old that were breast- or bottle- 

 fed and it also shows, by the addition of the deaths occurring within the 

 year at the several age periods, the total number of breast- and bottle-fed 

 babies and the percentages, corrected by these deaths, for the various 

 age periods. Table 119 shows the mortality of infants in the City of 

 Boston in 1911, by age, nativity of mother and by feeding. Table 120 

 shows the mortality of infants in the City of Boston in 1911, by month, 

 age and feeding. In this table the slight increase of deaths among breast- 

 fed babies during the summer months is strikingly contrasted with the 

 multifold increase in deaths among bottle-fed babies in July, August 

 and September. Table 121 is a comparison of the actual infant death 

 rates and estimated infant death rates, had all babies been breast-fed. 



"The conclusions that may be reached from the figures in this last table are 

 startling. For instance, had all the infants above 2 weeks of age been breast-fed 

 or equally well-fed, for there is reason to believe that bottle feeding may be so 

 safeguarded that it will be productive of no greater infant mortality than breast 

 feeding, there would have been but 1,253 infant deaths, whereas had all been 

 bottle-fed there would have been 4,352 deaths. The actual number of infant 

 deaths in 1911 was 2,245, giving an infant death rate of 127 per 1,000 births but 



