No. 34-] HEMIPTERA OF CONNECTICUT: MIRIDAE. 425 



the family Miridae, the writer has arrived at certain conclusions 

 regarding the relationships of the subfamilies. Perhaps the most 

 significant change is in the elevation of the subfamily Orthotylinae, 

 from next to the lowest subfamily where Reuter placed it, to a 

 place near the top of the series,' or near to the Capsinae and 

 Mirinae. When guided by the form of the arolia and the genital 

 structures, we are dealing with the most fundamental characters 

 yet proposed for classification of the subfamilies. On the basis 

 of these structures, it may readily be seen from the figures accom- 

 panying this paper, that the subfamily Orthotylinae is more closely 

 related to the Capsinae than any other subfamily. Reuter placed 

 the Orthotylinae next to the Phylinae because of general similarity, 

 chiefly on absence of pronotal collar, rather than relying on the 

 form of the arolia. The present writer not only finds a close 

 relationship between the erect, diverging and converging arolia 

 (Fig. 49), but also in the highly developed and specialized 

 character of the genitalia. 



J. Sahlberg (1920) and Bergroth (1922) find reason to recog- 

 nize in the Mirinae of Reuter, two subfamilies, the Capsinae and 

 Mirinae. Although both groups have erect arolia, divergent on 

 the apical half, there are still very good characters for recognizing 

 two subfamilies. Certainly the Mirinae are more primitive than 

 the Capsinae. In addition to the characters given in the subfamily 

 key, the Mirinae may be judged more primitive as indicated by: 

 (1) more generalized and little modified type of genitalia; (2) 

 few genera and each with comparatively few species, but several 

 of these are of world-wide distribution; (3) all the species breed 

 on grasses or related lower plants. The Capsinae, on the other 

 hand, are more specialized and stand higher for the following 

 reasons: (1) possess highly specialized genitalia; (2) numerous 

 genera and species but having limited distribution; (3) the numer- 

 ous species breed on a great variety of the higher plants, or plants 

 of comparatively recent origin, also many species have developed 

 predaceous habits. 



The phylogeny of the Mirid subfamilies does not present a 

 linear series of development, but more of a progression upward in 

 several directions, and this perhaps may best be represented by a 

 genealogical tree. A subfamily tree is appended (fig. 48) which 

 will express more clearly the relationships within the family, based 

 on the following characters which are listed in the order of their 

 relative importance: (1) arolia; (2) genital structures; (3) 

 biology; (4) modifications of the thorax. 



Key to subfamilies of Miridae. 



(Figure numbers refer only to Figure 49.) 



1. Arolia present, erect and prominent (fig. 49: 33-6o) ; or pseud- 

 arolia very prominent, usually with bristle-like arolia also present 

 (fig. 49. 13-20,22-24) 6 



