36 L- W. SACKETT 



The slight increase in time, and particularly in the mean 

 variation for the button and two combinations is, no doubt, 

 due to partial satiety of the animal as he approached the close 

 of the feeding period in each case. It should be noted that the 

 column showing the average of all does not indicate an increase 

 in time, and the average variation is only one second greater 

 than the preceding group. 



The last column of figures is represented graphically in plate 

 I by curves B and C. Curve B shows the diminishing time 

 used in operating various locking devices and combinations. 

 Curve C shows the average mean variation. It is clear that 

 improvement in the two respects goes on almost parallel. As 

 the animal reduces the time necessary to operate any partic- 

 ular device, or set of devices, he also becomes more regular in 

 his method and consequently shows less and less variation in 

 the time required. The only instance where the mean varia- 

 tion is greater than the average is in the first five trials at the 

 plug where the average time is 71.4 seconds and the mean varia- 

 tion is ±97.4 seconds. 



Little would be gained from combining curves of other ani- 

 mals with that of No. 11 to show a general learning curve of 

 the porcupines for the reason that results are, in many respects 

 not comparable. For the same reason it would be misleading 

 to combine curves of other animals observed by other experi- 

 menters in an effort to get a still more generalized composite 

 curve. Before such a combination would be at all illuminating 

 certain very obvious conditions must be kept uniform. 



1. Experimenters must agree on just what stage of the ani- 

 mal's attack shall be considered proper to start the time record. 

 Shall it be when the animal first starts toward the box? Or, 

 shall it be when he first attacks the particular device then 

 in use? 



2. There must be definite agreement as to when the animal 

 has opened the box. Shall the experimenter say that when the 

 door swings open the animal has succeeded, and disregard the 

 two, three or even 20 seconds which he wastes in useless effort 

 after the food is exposed? Or, shall the watch be stopped when 

 the animal ceases to work with the device and directs his activ- 

 ity toward an open box rather than toward a closed one? How 



