219 



The details previously given are collated in the following 

 table, in which the figures, when standing alone, represent per- 

 centages of individuals of the variety with bluish to yellowish 

 tibicP, instead of red. E.sti mated percentages are queried, and 

 dashes indicate lack of data or non-occurrence of species. In 

 the absence of precise data the proportion is expressed by words, 

 or. if unknown, is indicated by the plus mark ( + ). 



DISTRIIIl'TIOX AND I'ROroRTION OF MEI.ANOIM.rs II A VIN(i BLUISH TO YKLM IWISfl IHN'DTIBL^. 



Species 



t'emur-i'ul)ruii 



atlanis 



minor 



angustipennis 



and 

 coccineipes 



flavidus 



packardii 



spretus 



femoratus 



and 

 bivittatus 



East. 

 Sand(?) 



+ 

 + (N.H. 

 &Mass.) 



The variation in tibial color has no marked relation to sex. 

 In the small series of M. minor at hand the tibiiB are bluish, or 

 in the female sometimes red; in M. aiigHstipcunis from Wauke- 

 gan two of every three males have red tibi.e (21 of 83). and two 

 of every three females hav.e them bluish ( 20 of 31 1. 



This tibial variation belongs to what B.iteson calls discon- 

 tinuous variation, the transition from red to bluish in a series 

 of specimens being more or less sudden. Specimens at the 

 point of change show one color basally and the other apically. 

 This variation is attributed by Cockerell ('89) to an influence 

 promoting or arresting a metabolism of pigment. The order of 



