REHN AND HEBARD 211 



lation of same a little less than 90°, caudal margin rather irregu- 

 larly convex to the rather weak humeral sinus, convex callosity 

 moderately but not decidedly broad. As in attenuatus, the 

 ventral margin and particularly the ventro-caudal angle is 

 sharply but narrowly curved outward. 



Macropterism is found in but four females, all from Cumberland 

 Island, Georgia, in the series of over one hundred and sixty- 

 eight specimens before us. 



When compared with spartinae, the male cerci are found to 

 be very similar but more attenuate and slightly irregular in out- 

 line, this irregularity giving the organs the appearance of being 

 a little malformed. 



The ovipositor is normally rather broad and approximately 

 straight in direction with a weak open-sigmoid curvature. In 

 the series of eight specimens from Cedar Keys, Florida, we 

 find this type in three, and a distinctly thought not strongly 

 upward curved type in five, the ovipositor in these being broader 

 and showing an even greater curvature than is normal in spar- 

 tinae. The development of two distinct types of ovipositor in 

 the same species at one locality is a problem which we have also 

 encountered in Orchelimum concinnum. The females here con- 

 sidered belong without the slightest doubt to the same species 

 and the cause of this varied development is yet highly prob- 

 lematical. Elsewhere in the species of the genus some individual 

 variation naturally occurs in degree of curvature and heaviness 

 of the ovipositor, but the appearance here of two distinct tj^pes, 

 elsew^here of decided importance and value as specific characters, 

 is very surprising. Differentiation in method of oviposition 

 and selection of certain different plants for this purpose has 

 probably been a major factor in the development of the different 

 types of ovipositor now to be found in various species of the 

 genus, but when two distinct types are found in the same species 

 it would lead one to suppose them to be the result of these same 

 factors. The difficulty is that with such development the 

 different forms have as a rule developed into distinct species, 

 which in the present instance is not the case. The ovipositor 

 length is as follows: Ventnor, New Jersey, 10.7-12; Ocean 

 View, New Jersey, 11.1-12.8; Ocean View, Virginia, 12.7-13.8; 

 Wrightsville, North Carolina, 11.4-12; Cumberland Island, 



TRANS. AM. EXT. SOC, XLI. 



