from the internal study of the families separately and from the curves of correlated relationships in the first and second degree. Second, the belief of the author that environment would not cause the great names to be associated in blood.
If a great king succeeds in building up an extended empire, it does not appear to the writer that his son would have an easier task in becoming famous as a great governor. The importance which Prussia assumed under the Great Elector did not make his son, Frederick I., rank any higher than 3 in our scale. Still the genius has always been properly perpetuated somewhere, in some of the descendants. Also the times have continually 'called for great men.' Never did a dying country call more urgently than Spain in her last three centuries; yet none appeared. Italy had to wait fifty years for Cavour, Garibaldi and Victor Emanuel. England could not get a good Stuart, but in a descendant of William the Silent she found a hero in William III.
When we come to analyzing the moral qualities we find more difficulty than in the mental. First, because good parentage would tend to bring a better environment; and, second, because the curves of distribution, though in general entirely compatible with hereditary influence, are less perfect. Still it has come to be the belief of the author that even on the moral side heredity is more important than surroundings for the following reasons: First let us consider the curves of relationship. It will be seen (Plate II.) that the men and women in the high grades have some three times as many relations of their own kind as the bad characters have in the way of relationship to these best ranks. The degenerates (grades 1, 2, 3,) have between two or three times as many relations of their own ilk as the best have in these lower grades, while the best have actually more in their own 9, 10 grades than they have in the three lower grades put together. A second reason is drawn from a count of the different grades relative to the period in which they lived. We might expect that in the old days, when the standard of morality was rough, lawlessness and licentiousness would be found in a greater percentage than during more recent times. Three divisions have been made.
The period prior to the year 1(300 is here called 'old'; from 1600 to 1800, 'middle'; from 1800 onward, 'recent.' It will be seen in the chart (Plate III.) that the proportionate distinction of characters according to this method arranges them in each rank, almost perfectly according to the law of 'deviation from an average,' each group old, middle and recent, falling off equally from the mediocrities, so that when we attempt to make curves for old, middle and recent, according to the percentage in each rank, we get no curves at all, but lines almost flat. The only irregularities are at the edges 1, 2 and 9, 10 grades,
