60 THOMAS WILLIAM BROKBANK 



the redintegration of the rat which had the more poorly 

 established integrations* the dominant " error " would be 

 likely to occur first, and persist the longest. As with 

 the results on the maze it would be well to repeat that, 

 although innate constitution may bear some influence, 

 yet the data at hand on the inclined plane point to the 

 second explanation as more satisfactory, and more open 

 to proof than the intangible hypothesis of the first. 



In experiment on the inclined plane there are numer- 

 ous cases, not found frequently on the maze, where the 

 first trial or trials of redintegration were perfect and a 

 number of subsequent trials were imperfect; e.g., in the 

 records of rats 7, 9, 10, and 13. Had the present experi- 

 ment been conducted according to the norm which holds 

 that the first trial is " pure " retention and evidently 

 adequate to ascertain how well a habit is retained, the 

 individuals mentioned would certainly have been credited 

 with perfect retention. And further, according to this 

 norm, if the first trial after the retention period was perfect, 

 it would logically and theoretically follow that, with every- 

 thing else equal, the subsequent trials would also be per- 

 fect, in accordance with the theories of recency, frequency, 

 and repetition or intensity. The subsequent trials in the 

 cases cited were not perfect; and thus it may be inferred 

 that the capacity of the individual organism to retain 

 integrations may not be completely known in the first 

 trial after retention, but only when a number of successive 

 perfect trials show conclusively that the habit has been 

 acquired. It is thus important to remember that all 

 " errors " and imperfect integrations of the redintegration 

 series of trials throw light on how well the habit has been 

 learned in the first place and how well it was retained in 

 the second. Or it may be said further that the problem 

 is so difficult that it is impossible to get six consecutive 

 trials in redintegration since it is so difficult to get these 

 in learning; but all these rats made six consecutive trials 

 perfect in order to attain the norm. 



From the data of both litters the dominant " error " 

 was found, as stated in the previous records on the plane, 

 at the point in the problem where the rat must establish 



