46 



BULLETIN 114, UNITED STATES NATIONAL MUSEUM. 



Anniston, Alabama, collected in 1919 by E. R. Dunn, is practically 

 transitional. This specimen shows the crossbands to be almost 

 obsolete, but where they can be made out they seem to be reduced 

 in number. The alternating light and dark areas on the sides of the 

 belly are about 33 in number — a getulus character, but in general 

 appearance it is nearer to niger. This enables us to determine 

 fairly closely the limits of the ranges of these two forms in the South 

 and the region where intergrades may be looked for. 



That niger bridges the gap between TioTbroold and getulus there is 

 hardly room for doubt. Also that on the whole it is more closely 

 allied to JioTbrooki is evident from consideration of the characteristics 

 of all three. And from consideration of the latter and the geographic 

 relationships, it becomes most logical to regard niger as a derivative 

 of JioTbrooki. This makes it necessary to consider niger as ancestral 

 to getulus. We may therefore express the probable relationships of 

 these three forms as follows: 



TiolhrooJci—^niger— ^getulus. 



Further evidence for this view of their relationships will be found 

 in the summary of the getulus group. 



Scale formulae of niger. 



Formula. 



Male. 



Female. 



Totals. 



21-23-21-19 



1 



5 

 8 

 6 





1 



8 



14 



6 



21-19 



3 



6 



19-21-19 



19-21-19-17 



Total 





20 



9 



29 





