(g) July, 6 and 7, a fairly definite tendency to take the one next in order or, instead, to go directly to the right box. (h) July 10, direct first choices without approach to other boxes appeared for the first time on this date. For this problem, it proved impossible to establish and maintain uniform conditions of experimentation. Instead, because of the failure of the animal to improve and the tendency to discouragement, both punishment and reward had to be altered from time to time, and other and more radical changes were occasionally made in the experimental procedure. Below for the sake of condensed and consecutive presentation, the most important conditions from day to day are arranged in tabular form: CONDITIONS OF EXPERIMENT FROM DAY To DAY FOR PROBLEM 2 Date Punishment Reward May 17 ............. 20 sec. confinement ........ Food in right box for each (Aid after 10 trials) trial " 18 to 21 ........ 30 sec. confinement ........ Food (banana) in right box for each trial " 22 to June 2 .... Illness, no experiments June 3 ............... 15 sec. confinement ........ Food (banana) in right box for each trial " 4 ............... 30 " " ............. Food (banana) in right box for each trial " 5-10 ............ 60 " " ............. Beet, carrot and loquat, in addition to banana " 11 .............. 10 to 30 sec. confinement .. Beet, carrot and loquat, in addition to banana " 12 to 15 ........ 60 sec. confinement ........ Beet, carrot and loquat, in addition to banana " 16 .............. 60 " " ............. Banana and sweet corn--former preferred " 17 (1st series). 60 sec. confinement ........ Food (banana, as in early series) " 17 (2nd series). No confinement in wrong box; Food only for correct first but instead, return to choices starting point by way of alleys " 18 to 22 ........ No confinement in wrong box; Food only for correct first but instead, return to choices starting point by way of alleys