But to complete the presentation of the results reached by the Census Bureau on this point, it should be added that if the results for the last twenty-year period be analyzed by decades a different conclusion is indicated. Comparison of the rates of growth of southern negroes and southern whites for those two decades shows that the rate of increase for southern negroes, 1880 to 1890, was to that of southern whites as 55 is to 100, while in the decade from 1890 to 1900 it was as 68 to 100. I confess myself skeptical of the accuracy of these figures. It is difficult for me to accept results which show on their face that the rate of increase of southern whites east of the Mississippi River was less, 1890 to 1900, than it was 1880 to 1890, the rate falling from 19.1 to 18.7, while that for southern negroes in the same area was much greater in the second decade, the rate rising from 10.6 to 15.7. At the same time I see nothing better at present than to mark these figures as questionable and to suspend judgment until the results for 1910 are published. It may be that the increase among the negroes has been affected by the marked prosperity of the South in recent years and has been affected more conspicuously than the figures for the whites. With reference to _sex_ it may be noted that there is an excess of females among the negro population of the United States, while this is not true either of the Indians or of the native whites. Strangely enough, this excess of females is found even at the very earliest ages. It is a general rule that the number of male children born exceeds the number of female. Among 100 children born, on the average about 51 are male and 49 female. The scanty records of births in cities where the negroes constitute a considerable element of the population, show that in this respect the negroes conform to the rule. Yet negro children even at the very earliest ages, as enumerated by the census, show an excess of females over males. This is true of negro children under one month, and of each of the four other subdivisions of age under one year. Indeed it is true for every year of age up to nine. It may be noted that this anomaly appears for the first time in the figures for 1900. Whether it is due to the fact that that census first made the distinction between negro population and the total colored, including the Indians and Mongolians, I am unable to say.