No doubt many legends of the ancient world, though not really history, are myths which have arisen by reasoning on actual events, as definite as that which, some four years ago, was terrifying the peasant mind in North Germany, and especially in Posen. The report had spread far and wide that all Catholic children with black hair and blue eyes were to be sent out of the country, some said to Russia; while others declared that it was the King of Prussia who had been playing cards with the Sultan of Turkey, and had staked and lost 40,000 fair-haired, blue-eyed children; and there were Moors travelling about in covered carts to collect them; and the schoolmasters were helping, for they were to have five dollars for every child they handed over. For a time popular excitement was quite serious; the parents kept their children away from school and hid them, and when they appeared in the streets of the market town the little ones clung to them with terrified looks.... One schoolmaster, who evidently knew his people, assured the terrified parents that it was only the children with blue hair and green eyes that were wanted—an explanation that sent them home quite comforted. Observations of external characters, combined with precise measurements, have now been made on a large scale in most European countries, and these methods are adopted on anthropological expeditions. In this way a great deal of valuable material for study has been accumulated, but much work remains to be done in this direction. [Sidenote: =Methods of Dealing with Anthropometric Data.=] Not only have head, body, and limb measurements been recorded, but the device of an “index” has been adopted which gives the ratio between two measurements, as, for example, in the previously-mentioned cephalic index (p. 34). The averages or means of series of indices obtained from one people have been compared with those obtained from other peoples; but this method is misleading, as there is frequently a very considerable range in any given series, and a mean merely gives a colourless conception of racial types, the only value of which is a ready standard of comparison, which, however, is full of pitfalls. A further step in the advancement of anthropometric research was made when the extent and frequency of such deviations from the mean were recorded. At first this was done in a tabular manner by means of seriations; then curves were employed: a single peak was held to indicate purity of race, double peaks that two racial elements entered into the series measured, a broad peak or plateau was interpreted as being due to race fusion. Dr. C. S. Myers,[30] who has discussed these and other methods, points out the fallacies of this interpretation, saying: “There can be little doubt that most of the many-peaked curves owe their irregularity to the inadequate number of individual measurements which have been taken.”