considerably above that of Suffolk; and four other counties in that state have a rate practically identical with the Boston one. Now the European statistics prove conclusively that the city rate is from three to five times that of the surrounding country (Bertillon, page 55). To reconcile the discrepancy between these figures and ours, it must be assumed that some counterbalancing influences are at work here. Our foreign born and Catholic population have a divorce rate about one-fourth that of the rest of the country, and they are so massed in the cities as appreciably to affect the city rate. The theory that their presence masks the true influence of city life derives support from the figures of the state censuses; but those data are too inaccurate to be cited in detail.

§ 18. Distribution between Couples with Children and those without Children.

It is probable that the presence of children in a family diminishes the tendency to divorce. Can statistics test this theory and give any measure of the influence thus exerted ? If the number of childless couples, the number with children and also the number of divorces in each class, were known, then the tendency to divorce in each might be calculated, and the difference, if any, be found. A rough approximation to an answer may be obtained by combining some figures in the Report with others in the Massachusetts census of 1885 That state is, perhaps, the only one to report the number of childless wives and wives with children. The former are 17.56 per cent. of the whole number of married women. There is only one state, New Jersey, in which the court records indicate with fair completeness whether the parties divorced had or had not children. In that state the number reported as unknown was only 2.6 per cent. Three other states, Minnesota, Maryland and Colorado, report as unknown, respectively, 8 per cent., 14 per cent., 14 per cent. All the rest have records so incomplete as to be worthless for our purpose. By a combination of these figures with the estimated number of couples

for 1875 and the Massachusetts ratio of childless wives to the whole, the following results are obtained:

The errors involved in applying to a frontier state like Colorado a Massachusetts birth rate and an inaccurate ratio of couples to population, determined from a few eastern states, may account for the variation shown. On the whole it is fair to conclude, notwithstanding, that childless marriages are between three and four times as likely to end in divorce as marriages with children. The only foreign census cited by M. Bertillon (p. 135), as giving the same datum as that of Massachusetts, is the French census of 1856. The results obtained from that confirm the foregoing. The rate at which separations were demanded was 3.6 times as great among childless couples as among others.

§ 19. Distribution of Divorces as granted to Husband or Wife.

Nearly two-thirds of the divorces (65.8 per cent.) were granted on demand of the wife. The party going into court is usually somewhat more innocent than the other. Therefore the husband more often destroys the marriage tie than the wife; certainly he is more likely to commit those overt acts on which a suit for divorce must usually be based.

Law is for the protection of the weak against the strong through the superior strength of numbers, i. e. the community. But weakness may be beyond the effective protection of law. A modicum of knowledge and courage is necessary, or law is of no avail. Seldom can the police effectively restrain parental

abuse, and the submissiveness of negroes has foiled many an effort at interference on their behalf. In some parts of the country and some classes of society, a wife's relation to a husband is that of a child to a parent, or a servant to a master, much more than of an equal to an equal. Public opinion, her early training, dread of notoriety and the absence of all means of support apart from her husband, compel her to endure wrongs and ill-treatment to which the ear of the judge would be open. The purpose of the divorce law is foiled. Hence the number of divorces decreed is no more a test of the wrongs endured by wives, than the number of parents punished for cruelty to children is a criterion of the abuse they suffer. Divorces to wives measure their resistance, not their burdens. Southern wives are probably more resigned and submissive than northern, and, in accordance with this fact, the Report shows that they are less inclined to seek a divorce. In seven southern states, less than half the divorces are granted to the wife, and in the whole sixteen the ratio is only 55.4, while in thirty northern and western states and territories 69.2 per cent. are granted to women. If the number of divorces granted to the husband had been unchanged, 112,540, but those granted to wives had remained through the twenty years at the southern ratio, 55.4 per cent. of the whole, the total number would have been 251,776, instead of 318,716, and the number granted to northern women 95,240, instead of 172,183. The differences between the various states and territories in this regard may best be indicated by the following table:

TABLE VIII.

Percentage of Divorces to Wife in States and Territories.

The average length of marriages ending in a divorce granted to the wife is somewhat greater than of those where the divorce is granted to the husband. In other words, wives show a little more patience and long-suffering than husbands. The difference, however, is but trifling, three-tenths of a year, and even this is not found all over the country. Of the eighteen states and territories west of the Mississippi, all but Missouri, Nevada, and Utah have a shorter average interval between marriage and divorce in cases where the wife is the complainant. On the other hand, of the twenty-seven states east of that river, including Louisiana, Minnesota and the District of Columbia, twenty-one show a longer duration of marriage when divorce is decreed to the wife. The six exceptions are three states in the southeast, Georgia, Florida and Alabama, two in the northwest, Michigan and Wisconsin, and.one in the northeast, Rhode Island. The Mississippi, then, divides the country into two belts, an eastern, where suffering wives show a little more

*From 1875 to 1879 hundreds of divorces for non-resident eastern parties were granted in Utah (compare § 28). Most of these, fully seven-tenths, were to the husband. Hence the true position of Utah is found by neglecting the figures for those four years, and computing the percentage from the other sixteen.

patience than husbands, and a western, where the reverse is true.

§ 20. Distribution among the Several States.

The divorce rates of the states given in the table below have been computed as follows. The number of divorces in 1870 and 1880 was determined by averaging the figures for the seven years, 1867-73 and 1877-83; this average was divided in each case by the estimated number of married couples in the state for that year, and the decimal thus obtained was multiplied by 100,000. The result obviously expresses the number of divorces annually to 100,000 couples.