68 



DAY AND BENTLEY 



Table I records the first fifteen reversals for eight individuals. 

 Thus D turned in the tube twenty-five times before it brought 

 its axis around more than 90° and swam to the opposite end of 

 the pool. The next reversal came at the twenty-ninth turn, 

 and so on. The figures at the bottom of the table show" that 

 the first two reversals required, on the average, the greatest 

 number of abortive, partial turns (22.6 and 15.5). After that, 

 the averages fall into three groups: 3-4, 5-8, 9-15. The mean 

 variation (last horizontal line) follows the same course. The 

 averages and mean variation for the four groups are: 19.0 ±3.6, 

 8.i±i.5, 5-5±-3. 3-5±-5- 



TABLE I 



Number of Trials Necessary to Reversal for the First Fifteen Reversals 



Table II gives the total times, in seconds, consumed between 

 successive reversals. As an expression of change in behavior, 

 it is less satisfactory than Table I, because the subjects some- 

 times became inactive between " trials;" that is, they " wasted 

 time," and since activity tended to diminish with the lapse of 

 time, the periods of quiescence in the latter part of the table 

 mask the actual increase of facility in reversing. A comparison 

 of the two tables, line by line, will make this fact evident. The 



