The time of this trip was sixteen minutes. The times of some of the other trials were as follows:
| 10th trip | 4 | minutes. | |
| 15th" | 6 | " | |
| 20th" | 4 | " | 5 minutes |
| 25th" | 3 | " | |
| 30th" | 3 | " | 20 seconds |
| 35th" | 2 | " | 45" |
| 40th" | 4 | " | 20" |
| 45th" | 7 | " | |
| 50th" | 4 | " | 10" |
The route for the thirtieth trip was, as Fig. 6 indicates, almost direct.
The times of these experiments are generally longer than those of the first series because the inclines consumed considerable time.
During the formation of the habit of crawling up incline 3 and sliding down incline 4 a very interesting modification of the action occurred, namely, the shortening of the path to the sand-box by crawling over the edge of incline 4. At first the animal, after climbing up 3, would slide all the way to the bottom of 4 and would then turn toward the nest. Soon, however, it began making the turn toward the nest before reaching the bottom, thus throwing itself over the edge of 4. The turn was made earlier and earlier on 4, until finally it got to crawling over as soon as it reached the top of 3, or M. It always turned itself over the edge carefully, and landed, after a fall of four inches, usually on its head or back. By this process the path was shortened eight or ten inches. This action is a splendid illustration of the way in which an advantageous habit may grow by accretion, as it were, until it seems as if it must have been the result of reasoning. Some would, no doubt, hold that in this case the turtle chose the direct path because of inferences from judgments. Although this may be true, there is surely a sufficient explanation of the habit, as we have come to know it, in the profiting by chance experience. No one would say that the nest was at first found by inferences. It was reached because of the animal's impulse to move about, to seek escape or hiding. Had the turtle stopped to judge and draw inferences as to the way to escape, instead of persistently moving from place to place, it would probably be in the pen yet. No; the wandering impulse led by chance to the finding of satisfaction, turtle pleasure, in the nest. Because of this satisfaction, the action was impressed on the vital mechanism, so that there was a tendency (the beginning of a habit) toward repetition of it. Had the action failed to give satisfaction, the probability of its being repeated would have been
