DISCRIMINATIVE ABILITY OF THE TURTLE 25 



section of the present work illustrates definitely the effect of 

 such experience upon the animals concerned. This is found in 

 the pronounced preference of turtles Nos. 8 and 9 for black 

 boxes as the result of having participated in a former experi- 

 ment in which the boxes were painted black to emphasize a 

 white pattern. Had this former experience not been known, 

 and had the turtles in the black-white tests been fed in black 

 boxes without preference tests, the results would probably have 

 given an erroneous idea of these turtles' mental ability. This 

 is, of course, a rather special case, but it serves as a reminder 

 that in dealing with such illusive qualities as mental character- 

 istics the utmost caution must be taken if errors of judgment 

 are to be avoided. On these grounds the results obtained with 

 Nos. I and 11 in the black-white tests and also their memory 

 tests are open to question since no preference tests were here 

 taken. 



In discussing No. lo's record for discrimination between lines 

 of different width it was suggested that the relative ease with 

 which it discriminated between lines 2 mm. and 4 mm., and 

 2 mm. and 3 mm. in width might partly be explained by the 

 fact that immediately before these tests it had passed through 

 a large number of tests of a somewhat similar nature; in other 

 words, that its former experience had to a certain extent pre- 

 pared its mind to appreciate the latter problem. This is merely 

 a suggestion and no definite claims are made that the animal's 

 mind was more acutely receptive through having previously 

 solved somew^hat similar problems. There is far from being 

 sufficient evidence upon which to base statements of this turtle's 

 abilitv to profit by such educational processes. 



The relation of time to successful trials. My work with Chry- 

 semys shows conclusively that for this animal any calculation 

 of mental ability based upon the time taken by the turtle on 

 its trips from entrance room to food or electric box can give no 

 accurate measure of its progress in learning. This is particu- 

 larly true if several individuals are classed together and com- 

 pared, and it also applies to a comparison of the performances 

 of any one individual if compared at different periods during 

 the X)rogress of an experiment. Not only do different individuals 

 vary greatly in the speed with which they traverse the runway 

 and make choice of the boxes, but any individual's speed may 



