204 PROCEEDINGS OF THE ACADEMY OF 



NN. A discal areolet. 



0. Antenme of the tf much longer than the body ; those of the 9 short ; 

 second joint of the palpi longer than the third. Arrhenica nob. 



00. Antennae ^ of $ and of about the same length ; both much shorter 

 than the body ; second and third joints of the palpi about the same 

 length. Eriocera Macq. 



MM. Antennae 13 (sometimes apparently 12 ?) jointed ; mediastinal cross- 

 vein, far remote from the tip of the mediastinal vein, anterior to the 

 origin of the petiole ; wings like Walker, Ins. Brit. Dipt. iii. tab. 

 xxx. f. 7, (having two cross-veins between the upper branch of the 

 radial fork and the subcostal vein.) Dicranota Zett. 



MMM. Antenme 16 (or 17) jointed. 



Q. Mediastinal cross-vein far remote from the tip of the mediastinal vein, 

 anterior to the origin of the petiole. 



R. Wings pubescent. Ula Halid. 



Dn . ., ( Amalopis Halid. 



RR. Wings glabrous. j Fedicia Zatr. 



QQ. Mediastinal cross-vein close by the tip of the mediastinal vein, 

 or not far remote from it, and always posterior to the origin of the 

 petiole. Limnophila nob. 



Limnobia. nob. 



(L i m n o b i a Meig , ex parte, inclusive of R h i p i d i a Meig., G e r a n o- 

 myia Hal , and Dicranomyia Steph.) 



One radial area. Antennae 14-joiated, (sometimes apparently 15-jointed, 

 the cylindrical prolongation of the last joint appearing in some species as an 

 independent joint). Ungues armed with teeth ou the under side. Tibia? with- 

 out spurs at the tips. Pulvilli indistinct or none. 



The proboscis is always elongated, cylindrical, although generally shorter 

 than the head; in Geranomyia alone (1st Section) it reaches extraor- 

 dinary proportions. The $ genitals vary in structure in the different sections 

 of this genus; but all their forms have one character in common, and that is, 

 the great development of the anal style, situated immediately under the forceps. 

 (This organ is designated by the letter e on the figures 1, 3, 4, 5, 6, etc., of the 

 plate.) 



The group thus characterized, although very natural, comprises various forms 

 on which new subdivisions may be established. But there is no reason to 

 isolate some of them, leaving the others under the old generic denomination. 

 The genus, as defined above, is so compact, that either it must remain in its 

 integrity as a group of higher order, or the whole of it must be subdivided into 

 subordinate groups. This rule has not been observed in former works. 



The genus Rhipidia has been established on a European species with 

 pectinated antennae; the genus Geranomyia on several species with a 

 long proboscis ; but both R h i p i d i a and Geranomyia have more affinity 

 to some species which remained in the genus L imnobia (especially to those 

 which form my 3d Section, Dicranomyia Steph.,) than these species have 

 to their neighbors of the same genus, (to the L i m n o b i ae of my 5th Section.) 



The pectinated or moniliform antennas of R h i p i d i a are a character of 

 altogether secondary importance ; that is they do not indicate a corresponding 

 modification in the rest of the organization. The inconstance of this character 

 is alone sufficient to prove its secondary value : in Rhipidia maculata the 

 antennae are pectinated in the male and moniliform in the female; in Rhipidia 

 domestica n. sp. they are moniliform in both sexes. Such a character may 

 be used in the classification, but it must be subordinate to more important 

 characters. 



In Geranomyia likewise the length of the rostrum, although a very good 

 generic character, has apparently but little influence on the rest of the organi- 



[Aug. 



