TABLE X. FORM ASSOCIATION. DISTRIBUTION OF RIGHT CHOICES

Choices from series 1 to 5 in the case of Tri., Sq., Hex., and Cyl. boxes
Food in Tri. Food in Sq. Food in Hex. Food in Cyl.
Animals1,2,3,4,5, 1,2,3,4,5, 1,2,3,4,5, 1,2,3,4,5,
(U)20213 12222 21233 11213
(V)11232 21122 01232 12211
(W)12122 21212 12223 12232
(X)12332 02231 12211 11232
(Y)12323 22232 11213 02223
(Z)12221 02332 12242 12312
Total,79131313 710121311 69121414 510131113

It will be seen that each animal chose the right box oftener than any other one box, but not oftener than all of them; also that there was a small increase in the number of right choices from series to series. No one of the four forms seemed better discriminated than the others if we may judge from the practical equality of right choices made in each case (55, 53, 55, 52) or from the similar increase in number of right choices from series to series; the hexagonal and cylindrical boxes received fewer choices in the first series than did the triangular and square, but this was exactly counterbalanced in the last series. The triangular box was more often confused with hexagonal and square, and the square with triangular and oblong, than with the others. For the hexagonal box the cylindrical was more frequently mistaken than were the other forms, especially the oblong; and with the cylindrical the elliptical was more frequently confused than were the others, especially the oblong. In this series of tests nothing new as regards general behavior or method of learning was observed.