MODIFICATION OF COI.OR. 

 Table ioo. — General color of beetles used. 



2IX 



Class 



6 





11 



12 



13 



u 



15 



i6 



I? 



Parents 



Control 



Experiment. . 



P.ct. 



I 



P.ct. 



I 



P.ct. 

 3 



I 



2 



P.ct. 

 4 

 5 

 4 



P.ct. 



7 

 8 



9 



P.ct. 

 i6 

 i8 

 14 



P.ct. 

 50 

 46 

 47 



20 

 22 



20 



2 



In experiments where an average increase in temperature of more than 

 12° C. was used the mortaHty was too high for the observations to be of any 

 value. Experiments conducted with lowered temperatures gave results like 

 those in increased temperatures. In fact, in all the experiments with signati- 

 collis the same result was obtained — that there was no modal shifting of the 

 species as the result of increased or decreased temperature. This is in close 

 accord with my observations of this species in nature, where it is apparently 

 extremely stable and conservative, and varies as little as in experiment. 



MoisTURK Experiments. 



L. signaticolHs was also experimented upon with deviations in relative hu- 

 midity similar to those used with decemlineata, but without any resulting 

 color modification. 



In signaticolHs is found a species which differs greatly from decemlineata in 

 that it does not respond to stimuli by altered coloration. Pure somatic varia- 

 tion in signaticolHs is therefore very limited and of a low degree, a fact that 

 has already been arrived at in the study of variations. We are not to conclude 

 from this that signaticolHs does not vary, because it gives rise to most inter- 

 esting modifications. These, however, are germinal, and not somatic. But, 

 as far as purely somatic modifications of color are concerned, signaticolHs is 

 highly conservative, and a striking contrast to decemlineata. 



MODIFICATIONS OF C0I.0R IN 0THI:r SPECIES OF IvEPTlNOTARSA. 



Experiments with the environmental factors used with L. decemUneata and 

 signaticolHs have to some extent been tried upon undecimlineata, dilecta, via- 

 lescens, oblongata, juncta, rubicunda, and multitceniata. From these experi- 

 ments, the data of which it seems unnecessary to give, I find that species like 

 multitmiiata and iindecimlineata, which are variable in nature, are likewise 

 variable in experiment ; while those which are stable in nature are also stable 

 in experiment. However, this rule applies only to the variations produced in 

 the organism during ontogeny — that is, after fertilization — by the action of 

 such external stimuli as those which are used in the production of the varia- 

 tions which we usually call fluctuating, and which are largely the cause of the 

 high place and geographical variability in adults. 



