The first Boston table is based upon about equal numbers of boys and girls, and children of Irish and American parentage greatly predominate; there are 21 Germans, and 19 are divided between 8 other nationalities. 14 per cent. of all examined did not know their age; 6 per cent. were four, 37 per cent. were five, 25 per cent. were six, 12 per cent. were seven, and 2 per cent. were eight years old. The returns were carefully tabulated to determine the influence of age, which seems surprisingly unpronounced, indicating, so far as the small numbers go, a slight value of age per se as an index of ripeness for school.

In the second table, which is based on Boston children, only columns 2 and 3 are based upon larger numbers and upon less carefully restricted selections from the aggregate returns. In

34 representative questions out of 49, the boys surpass the girls, as the German boys did in 75 per cent. of the Berlin questions. The girls excel in knowledge of the parts of the body, home and family life, rainbows, in conception of square, circle, and triangle, but not in that of cube, sphere, and pyramid, which is harder and later. Their stories are more imaginative, while their knowledge of things outward and remote, their power to sing and articulate correctly from dicta- · tion, their acquaintance with number and animals, is distinctly less than that of the boys. The Berlin report infers that the more common, near, or easy a notion is, the more likely are the girls to excel the boys, and vice versa. Save possibly in the knowledge of the parts of the body, our returns do not indicate difference between the sexes. Boys do seem, however, more likely than girls to be ignorant of common things right about them, of which knowledge is wont to be assumed. Column 4 shows that the Irish children tested were behind others on nearly all topics. The Irish girls decidedly outrank the Irish boys, the advantage to the sex being outweighed by the wider knowledge of the boys of other nationalities. Whether, however, the fiveand six-year-old Irish boys are not after all so constituted as to surpass their precocious American playmates later in school or adult life, as since Sigismund, many think "slow" children generally do, is one of the most serious questions for the philosophical educator. Column 6 shows in a striking way the advantage of the kindergarten children, without regard to nationality, over all others. Most of the latter tested were from the charity kindergartens, so that superior intelligence of home surroundings can hardly be assumed. Many of them had attended kindergarten but a short time, and the questions were so ordered that the questioners who had a special interest in the kindergarten should not know till near the end of the tests whether or not the children had

ever attended it. On the other hand, a somewhat larger proportion of the children from the kindergarten had been in the country. Yet, on the whole, we seem to have here an illustration of the law that we really see not what is near or impresses the retina, but what the attention is called and held to, and what interests are awakened and words found for. Of nearly thirty primary teachers questioned as to the difference between children from kindergartens and others, four saw no difference, and all the rest thought them better fitted for school work, instancing superior use of language, skill with the hand and slate, quickness, power of observation, singing, number, love of work, neatness, politeness, freedom from the benumbing school bashfulness, or power to draw from dictation. Many thought them at first more restless and talkative. There are many other details and more or less probable inferences, but the above are the chief. The work was laborious, involving about fifty thousand items in all. These results are, it is believed, to be in some degree the first opening of a field which should be specialized, and in which single concept groups should be subjected to more detailed study with larger numbers of children. One difficulty is to get essential points to test for. If these are not characteristic and typical, all such work is worthless. We believe that not only practical educational conclusions of great scope and importance may be based on or illustrated by such results, but, though many sources of inaccuracy may limit their value, that they are of great importance for anthropology and psychology. It is characteristic of an educated man, says Aristotle in substance, not to require a degree of scientific exactness on any subject greater than that which the subject admits. As scientific methods advance, not only are increasingly complex matters subjected to them, but probabilities (which guide nearly all our acts) more and more remote from mathematical certainty are valued.