300 CARL W. BOCK 



to the then resulting group as now attaches to the 22 group, 

 and since for present purposes analysis must cease at some point, 

 it might as well be at the 22 group as elsewhere. The impor- 

 tance that attaches to the 22 group, as it stands at present for the 

 reasons cited, does not depend upon the fact whether the 22 

 group is or is not an ultimate functional unit, if such exist; it 

 depends solely on the fact that the 22 group is a functionial entity 

 and that it is related to other groups in very definite ways. For 

 that reason the question as to its composition will be left open for 

 the present with this additional remark, that it is without doubt, 

 or was at some time a complex activity, compounded of smaller 

 units whose precise nature is at this time indeterminable. 



6. The coefficient y. The general series (6) contains a co- 

 efficient y which represents a class of groups whose description 

 and analysis have not been possible thus far. It is the purpose 

 of this section to consider this class of groups and to this end 

 the data of observer S will be studied (table 3). 



The question must first be raised and answered on what as- 

 sumptions it is possible to attempt a description and definition 

 of a general series of groups such as series (6) by means of an 

 appeal to the data of several different observers, defining cer- 

 tain of the coefficients of this series on the basis of the groups 

 of one observer and others on that of different observers. The 

 question contains the implication that different observers beat 

 si ch radically different groups or classes of groups that no gen- 

 eralizations are possible, or otherwise it ought to be possible to 

 completely define every coefficient of the general series (6) by 

 the data of a single observer, such as R. 



It is not the same to say that different observers beat differ- 

 ent groups and different observers beat different classes of groups. 

 Different observers do beat different groups, though they often 

 also beat similar or even identical groups. On probability 

 grounds this is to be expected considering the large number of 

 possible observers. But the statement that observers beat dif- 

 ferent classes of groups contains the contrary to fact implica- 

 tion that members of the same species differ very widely in struc- 

 ture and function, i.e., that they differ more than they are alike. 



