1916.] NATURAL SCIENCES OF PHILADELPHIA. 137 



shown us that coarctatus is merely this southern bolteri somewhat 

 modified from the typical more northern material, but so slightly 

 that there now appears no valid reason for retaining the name, even 

 in a racial sense. Curiously enough, however, it is in south and 

 central Florida only that long-pronotum individuals of the bolteri 

 type have been taken. Hancock's variabilis we find is merely a very 

 faintly emphasized type of femoratus, certainly too poorly defined 

 to be entitled to a name. In the femoratus type in south and central 

 Florida we find the nearest approach seen in that species to the 

 bolteri series, the two lines seeming to vary toward one another in 

 'that area in a fashion not found in our series from northern Florida 

 northward. This fluctuation apparently expresses a certain amount 

 of instability of the forms in that region, a probable breaking up 

 along other lines being under way, but .as far as taxonomy is con- 

 cerned it is undesirable, and in fact impossible, to attempt to express 

 or define the minute shades of difference seen, although they can be 

 appreciated in certain specimens. From the sum total of specimens 

 examined it is evident that some few specimens of " coarctatus" and 

 u variabilis" stand out from bolteri and femoratus, respectively, with 

 some degree of distinction, but the vast majority, into which the 

 distinctive specimens blend, cannot be separated unless one sorts 

 by localities. In consequence of the fluctuations of these Florida 

 specimens, the construction of a key separating femoratus and bolteri 

 has not been easy. A character which we suggested as diagnostic 

 in 1912, 29 i.e., subscutellate or regularly diverging frontal forks, we 

 now find to be unstable in both forms, although diagnostic in the 

 greater proportion of the specimens. In consequence the present 

 key for separating femoratus and bolteri is almost entirely comparative, 

 but we feel it will serve the purpose of separating two species which 

 over by far the greater portion of their ranges are clearly defined. 



Of Blatchley's hancocki we now have, through the kindness of the 

 author, a topotype (9 ; Knox County, Indiana, VII, 1, 1913, W. S. 

 Blatchley) before us. This specimen shows that the species is 

 inseparable from A 7 , bolteri, the individual in hand being indistin- 

 guishable from east coast material, of which, it will be seen, we 

 have a sufficient quantity to realize the extent of intra-specific 

 variation. 



Quite apart from the relationship of the previously known forms 

 of this genus from the southeastern States, we find in the present 

 series five specimens referable to this genus which are quite note- 



29 Proc. Acad. Nat. Sci. Phila., p. 244, (1912). 



