54 University of Michigan 



dant. For some reason this si^ecies was not found in any of 

 the western locahties where collections were made. 



These specimens belong to the typical race, fasciatus fasci- 

 atus, but they average somewhat smaller and darker than a 

 similar series of Michigan specimens. No specimens of the 

 northwestern race, A^. fasciatus abortivus Caudell, were taken 

 in any of the localities studied, but Caudell has reported abor- 

 tivus from Portal, Burke Covmty, in the northwestern part of 

 the state close to the Canadian boundary, and just within the 

 Drift Prairie region. ^^ 



It may be of interest to note that the caudal femur-ovipos- 

 itor ratios of 24 female specimens, 12 from Devils Lake and 

 12 from the Turtle Mountains, when plotted according to the 

 scheme devised by Hebard,"''- are distributed as follows : With 

 few exceptions the plotted ratios lie close to the point of 

 junction of the areas designated as typical of the three races, 

 fasciatus, abortivus, and socius; a total of 16 fall within the 

 abortivus area, 5 just over the line separating abortivus from 

 socius, and 3 well within the fasciatus area. Hebard has 

 emphasized the fact that no single character can be relied 

 upon to distinguish these races, but that it is necessary to 

 consider the sum of all of them in order to arrive at correct 

 conclusions. 



Gryllns assimilis (Fabricius). — Devils Lake, July 9, 1919, 

 I juvenile female; July 19-Aug. 15, 1920, 26 males, 30 females; 

 Stump Lake, July 24, 1919, 4 juveniles; Sheyenne River, Nel- 

 son Co., July 25, 1919, I juvenile female; Pembina, Pembina 

 Co., July 12, 1921 (C. Thompson), i juvenile female; Botti- 

 neau, Aug. I, 1920, I juvenile female; Aug. 9, 1920 (A. H. 



•'*'' Determined by Morgan Hebard. 



31 Proc. U. S. Nat. Mus.. xxxiv, 1908, p. 81. 



^^Proc. Acad. Nat. Sci. Phil., Ixv, 1913, p. 4i3- 



