164 BULLETIN OF THE 



of the Tenth Census, but at the usual hour of adjournment it was 

 interrupted, to be resumed at the following meeting. 



The Society then adjourned. 



200th Meeting. May 14, 1881. 



The President in the Chair. 



Thirty-six members present. 



The minutes of the last two meetings were read and adopted. 



The first communication of the evening was the continuation by 

 Mr. J. S. Billings of his remarks upon 



MORTALITY STATISTICS OF THE TENTH CENSUS. 

 [Abstract. ] 



Mr. J. S. Billings described the methods used in the Tenth 

 Census to secure completeness and accuracy in the returns of mor- 

 tality. The Superintendent of the Census sought to secure the aid 

 of the physicians of the country, and for this purpose sent to each 

 a small blank book, each leaf of which was arranged to record the 

 facts connected with a single death. 70,306 such books were issued, 

 and 24,057 returned at the end of the census year. The data from 

 these books were compiled by causes of death, age, and sex, and 

 the slips were then used to complete the enumerator's schedules. 

 The total number of deaths reported from all sources for the census 

 year will be a little over 800,000, or about 16 per 1,000 of living 

 population, being an improvement in completeness over previous 

 censuses. The results of the attempt to record the number sick on 

 the day of the census are not very satisfactory, and it is feared they 

 will be too incomplete to be used. Taking the schedules for the 

 State of Rhode Island, which are believed to be the most complete, 

 it is found that the number reported sick on the 30th of June was 

 11.18 per 1,000 of the whole population. 



It is usual to estimate two years of sickness to each death, which 

 would make the number constantly sick range from 30 to 40 per 

 1,000. In the army for five years the proportion was 43 per 1,000. 



It seems probable that, while the proportion of sick shown by 

 the Rhode Island count is too low, it is more nearly correct than 

 any other data which we possess. 



