Analysis of Heterogeneity in Some Simplest Chaeacteks 211 



Examinations of the other species that have formed the material basis of this 

 portion of the investigation with reference to these simplest characters show 

 the same kind of data and the same general results as those set forth. All of the 

 species in the group have been gone over in this way and none have given results 

 in any wise different from those stated. Only in detail, which is of no general 

 interest, do the sets of observations differ. Spot c has been most frequently 

 presented in this account, because of all the pronotal characters it is least liable 

 to fusions and least variable. In all instances this " simplest character " shows 

 heterogeneity which, by definition, may be made fluctuating or mutative, quanti- 

 tative or qualitative, as one may wish. It is shown 

 that the simplest characters differ in amounts when 

 measured, change in different directions, and enter leo 

 into diverse combinations. They are always de- ,50 

 limited, are apparently never undelimited, and may 

 or may not respond to heterogenetic causes in 

 definite or indefinite reactions, so that these trivial ""^o 

 characters react individually in much the same way uo 

 that individuals and species do. To some this may ^^^ 

 mean that the individual as a whole so responds, 

 and the character reacts in the same way because '°° 

 correlated with the whole. To some extent this is 90 

 true, but not entirely so, because these areas have a go 

 specificity of behavior and an individuality of re- ^^ 

 action and stability that is astounding. 



In L. undecimlineata, in the same locations, c is *° 

 often wanting in a low percentage of the popula- 50 

 tion, and from these individuals a race in which c 40 

 never appears can be easily created. Of interest, 

 therefore, is the experience of crossing one of these 

 races (biotype) based upon a minute and trivial ^° 

 character with L. diversa or L. signaticollis, where to 

 c is never absent, and then to discover that Fj 

 extractives of signaticollis with c absent are ob- ^ 

 tained, and from them arise pure biotype lines 

 differentiated by this character. The presence and 

 absence relation is to be conceived of in this in- 

 stance as appljdng only to the one area c, but 

 equally precise behaviors have been observed in h, d, and in the absence of 



am 

 the a' + &'+ + +a+b group. The contrasting characters, minute as they are, 



pm 

 behave in the same precise manner as do larger and more generally distributed 

 characters. This can only mean that somehow in the gametes there are exact 

 and detailed conditionings of this character (spot c), and that in crossing there 

 is a precise recombination of the germinal agencies, such as to introduce, where 

 it was not before, a condition of the character found in another species. WHiile it 

 may well be true that whatever does produce the new condition in the c-less 

 extracted signaticollis of F, may be tied to or be a property of the whole, there 

 still remains the grave difficulty of accounting for so exact and trivial a mani- 



Fig. 28. — Statistical treat- 

 ment of values of left spot c 

 in L. diversa and its variety 

 rugosa at El Borrego near 

 Orizaba, Sierra de Escamela, 

 and El Riego. 



