'226 Priinar// and Secondunj Reduplication Series 



For three factors taking part in a reduplicatidii series, Bateson and 

 Punnett (1) suggest an octant arrangement. 8uch an octant arrange- 

 ment showing the possible course of the divisions in the formation of 

 reduplication series is given below. 



imn 



It is difficult to see any cause for differential treatment on such 

 a scheme. Trow's scheme, however, which involves one factor waiting 

 its turn to segregate until the other two have completed their redupli- 

 cation series, does from its very nature offer a jjossible explanation for 

 such differential treatment. Consequently further research showing 

 whether the course of events is best explained by the general hypothesis 

 or by the special hypothesis may throw light upcjn the question as to 

 whether the octant scheme or Trow's scheme is to be preferred as a 

 better picture of the process of segi'egation and the formation of 

 reduplication series. 



An interesting feature that becomes apparent on analysing the 

 observed fact by means of the general hypothesis, is the regularity of 

 the underlying phenomena. Such analysis shows that in all four 

 cases the two fundamental series of least intensity have their intensity 

 reduced when they become primary series, although the observed or 

 secondary series may be of a greater intensity. It is by no means 

 impossible that the same holds good for the third reduplication series, 

 as is certainly the case in the EBL x ebl and in the DfN x dFn 

 uiatings. 



Another possible regularity in the observed phenomena is the 

 reduction of two of the fundamental series to primary series of 

 identical intensity. The numbers are strongly in favour of this sug- 

 gestion in at least two of the cases. If such a relatitmship is shown 

 to be general, it may be due to the necessity for the divisions involved 

 in segregation and in the formation of reduplication series to be on 

 a symmetrical jilan. 



