19 



rising above the midrib until they approximate and a thin-walled 

 hollow pod is formed. The galls may be sought for during the 

 months from June to October. The larvae pupate in the ground, 

 the imagines emerging in about six days after pupation."* The 

 galls on the leaves are certainly more conspicuous than are the 

 effects of the American species in this country, in rose-houses. 

 The species winters in the pupal stage. 



HABITS OF THE AMERICAN SPECIES. 



I have never seen the larvae attacking the native wild roses, 

 even about Chicago, though, in the near vicinity of infested rose- 

 houses, I have found them in buds of a variety known as the 

 Bride, growing in the open air. The manner of hibernation in 

 our species is not well understood, but in the rose-houses there 

 is no injury done by the larvae from late October until late the 

 following May. A thorough inspection of large rose-houses 

 about Chicago, November 18-20, did not reveal a single larva or 

 adult, even where serious damage had been wrought a few 

 months before, and an inspection made on January 29, 1904, 

 also failed to reveal their presence. 



In the rose-houses, the larvae attack the Meteor, Wooton, 

 Bride, Madame Chatenay, La France, Ivory, and Golden Gate, 

 but are far more fatal to the first than to any other variety. 

 The insect first appears in such rose-houses as are old and more 

 or less open on account of cracks and crevices, or else in such 

 as are new and tight, requiring much ventilation. In the case 

 previously referred to, where only an alley separated infested 

 from uninfested premises, the uninfested houses were given the 

 least possible ventilation consistent with the growth of the 

 roses, while those infested were more generally ventilated. 

 Later on, these conditions were reversed, and the situation as to 

 insects changed also, the premises that had before escaped being 

 now overrun. 



An experienced foreman in the employ of a firm of exten- 

 sive rose-growers whose premises were infested with this midge, 

 informed me that the ins ect first attracted his attention by at- 



*"British Vegetable Galls," by Edward T. Connold, p. 206, Plate 87. 



