about half of the entire enrollment for the year in all cities and villages of 4,000 population and upward. The tables which follow show the distribution of pupils in the different grades, summarized by sex, for each of the two classes of cities and for the two combined. The figures showing the proportion of boys and of girls in the different grades indicate that the girls vary from grade to grade less than the boys. This, of course, is a confirmation of general knowledge as to the greater persistence of girls a fact due to the larger industrial opportunities which begin to call the boys from school at an early age. Nineteen per cent of all pupils enrolled were in the first grade, and the percentage declines to 5 at the eighth grade.

The closer enforcement of existing compulsory education laws, and their adoption in places which have not heretofore passed laws of this character, will undoubtedly from year to year change these percentages. Statistics of grade enrollment may, then, from year to year be expected to show less abrupt breaks between consecutive grades. Irregular attendance is responsible in large degree for failure of promotion, and is the most frequent cause of the "repeater" in the different grades; and his presence in the same grade for parts of two or three years, in each of which he is reenrolled, has a marked effect upon the statistics on this subject. The congenital dullards, the subnormal children, who are not promoted because of their mental infirmities, bear a very small proportion to the whole number of students enrolled.

In the German Empire, where compulsory laws have been long in force and are executed very strictly, the enrollment in the eighth year is about two-thirds of that in the first. It must be considered in this connection, however, that in the older countries, where population is not increasing so rapidly as here and social conditions are otherwise more stable, greater uniformity will naturally be expected.

It is to be remembered that the ratios here shown do not give a proper index of elimination, because of the natural increase in population, and consequently the number of native children entering the first grade now is greater than was the number of children of corresponding age eight years ago, when the present eighth grade entered school. As for the increase in population by immigration, it may be assumed that families are less disposed to emigrate to a new country when the heads of these families have reached middle age or beyond, and have children old enough to enter the higher elementary grades, than are younger couples, who either are newly wed or whose children are of ter der age. The contribution to the lower grades from this source is considerable, while that to the higher grades is almost negligible. Even with the older children, ignorance of the language

The figures of enrollment by grade in the States of Maryland, North Carolina, Tennessee, Utah, and Washington are taken from the printed school reports of these States. They are introduced to show the distribution when both rural and urban conditions are combined. Grade distribution in 386 cities of 8,000 population and upward.

Grade distribution in 366 cities and villages of 4,000 to 8,000 population.

Boys.

Combined grade distribution in 752 cities and villages of 4,000 population and upward.

Grade.

Per cent in each grade of total enrollment.

Boys

Boys.

Girls. and girls.

Grand total.. 197, 186 100.0 483, 180 100.0 500, 156 100.0 66, 339 100.0 179, 260 100. 01, 426, 121 100. 0

a For 1906-7.

For 1903-4.

c Less than one-tenth of 1 per cent.