*THE POPULAR SCIENCE MONTHLY.*

actually sets out to find a red tennis ball on a grassy lawn, his steps being plotted while he is making the search. The ball was placed in all cases so that it could be seen from a distance of ten steps and at a uniform distance forty steps from the center stake, in order that the tracings of different persons could be compared as to time and length of course traversed. Fig. 9.

+Starting point. |

— — 40 step circle |

——— Ideal curve. |

——— Path of T. P. H. 4 minutes. |

— · —""T.L.B. 7" |

. . . . . ""L.W.5" |

No attempt was made to do more than was necessary to obtain a suggestion which might serve to indicate the logic and method of the pigeons. In all only sixteen tracings were taken, and this number includes one experiment on a shepherd dog. Of the whole number, ten conform more or less closely to Prof. Story's curve. Three of the best of these are reproduced in Fig. 9. Two correspond evidently to the circular type (Fig. 6). These are given in Fig. 10, together with the only one which is in any degree rectangular, the "Yankee" type, and this was not made by a Yankee at all, but by an Irish boy twelve years of age. Three of the curves are hardly susceptible of logical classification.

A point of interest in this connection attaches to the dotted line in Fig. 9. This represents the path of L. W., the university carpenter, a man sixty-five years old, who had worked at his trade of straight lines and right angles for forty-five years. I had expected a typical rectangular curve. Instead he gave an ideal spiral. After finding the ball, however, he volunteered the following suggestive remark: "After I got started, I thought, if I was going to do it again, I would go at it on the square. I started out," he added, "before I thought." Thus instinctive logic won the day against forty-five years of special training.

If instinctive, however, this logic should be found almost as well developed in children as in adults. Accordingly, spicing the ball with candy or a small coin, several experiments were tried upon children from three to twelve years of age. In general the above statement was supported; but two of the number, aged respectively three and six, failed to show anything like the amount of search-logic possessed by most of the pigeons or the shepherd