284 CARL W. BOCK 



lute measure. Moreover this criterion would exclude from con- 

 sideration those groups whose frequencies fall below that of the 

 average of all groups and which include groups that are very sig- 

 nificant on the bases of other criteria. For instance the group 416 

 occurs but a single time; yet it can be predicted that if the 

 observer beats a group in the 400 's, it ought to be precisely the 

 416 and no other 400 group. The frequency of this group is given 

 by the fraction 1/303. 



As a matter of fact the terms significant and insignificant 

 have only a present convenience and no real justification at all 

 except relatively speaking. Organisms do not behave on the 

 chance basis; all their activities are significant, only some are for 

 present purposes more so than others. There are many factors 

 which tend to maximize the difficulties of a rigorous application 

 of the frequency criterion, of which but one can properly be dis- 

 cussed at this time. Assuming that the group 416 (vide table 2) 

 is significant for present purposes and of the same order of sig- 

 nificance as the 22 group or the 46 group of the same observer, 

 it might be asked how it happens that the latter groups are 

 beat so frequently and the former only once? The answer is 

 simple enough. The frequency criterion can only apply where 

 or when all groups have an equal chance of recurrence, which is 

 not in accord with either fact or theory in the present instance. 

 In the first place, the duration of the 416 group is relatively 

 much greater than the durations of the groups 22 and 46. In the 

 second, the 416 is a much longer group. Theoretically it ought 

 to be expected that animals or humans that have a relatively 

 large repertoire of groups, some short, others of medium size, and 

 still others that are long, will beat the medium sized groups 

 most frequently. This is of course on the assumption that the 

 organism in question has received no special training. A long 

 group, or one whose duration is long, has not the same chance 

 of recurrence, just because it takes longer for its performance, 

 and consequently the same measure does not apply for these 

 groups as apply for the shorter groups, and particularly for the 

 medium sized groups. 



