Discussion 187 



results. One cannot, at present, in most cases be confident about precise 

 quantitative changes in performance with age ; what one can usually say 

 with justification in a properly designed and controlled study is, "If I 

 find any change, any decline in the inevitably biased sample that I 

 have got, then a fortiori there would be a decline in a representative or 

 truly comparable sample." 



Verzdr-McDougall: Dr. Welford, how far does memory come into 

 your tests ? Our animal experiments are very crude in comparison with 

 your human experiments, but we have the advantage in using animals 

 (in our case, rats) that we can test groups of different ages and make 

 longitudinal experiments. We have, however, a great problem about how 

 to produce a motivation that is comparable in the young and the old 

 individual. In longitudinal tests, we have a certain safeguard in that we 

 test rats — young, middle-aged and old — up to five times during their 

 lifespan on the same multiple maze, and we find that some of the very 

 old rats tested for the fifth time during their life show perfect memory 

 of the maze. The motivation is hunger, in all cases. They have an 

 interval of two months and are then brought to the maze again for a 

 memory test, and we may suppose that these older rats, since they show 

 perfect recall, are as well-motivated as the young. An interesting point 

 in our experiments with the older animals is that while some old rats, 

 up to the very day of their death, showed perfect recall of the maze 

 (which is perhaps relevant to Dr. Olbrich's point about the possible 

 pathological condition), others which had also successfully learned it 

 several times in their life suddenly showed, at the age of about two years, 

 complete loss of memory. Although they appeared to be perfectly well 

 motivated (insofar as we timed their running through the maze and their 

 time was comparable to that of younger rats), these old rats were now 

 completely incapable of recalling or relearning the known task. 



Welford: There are, I think, really two kinds of memory loss which 

 may come with age. One is of long-term memories which have been 

 acquired a long time ago. The other is a much shorter term and more 

 ephemeral memory. Mrs. Verzar-McDougall's experiments deal mainly 

 with the first kind. Our experiments, when they involve memory at all, 

 are concerned with the second kind. This enters into many problem- 

 solving tasks, and lowering of capacity of short-term retention is prob- 

 ably one of the main causes of decline with age at performance at this 

 kind of task. 



