17 



out any clearly defined pattern to the coloration. The adult 

 flies afterwards reared from these larvae did not entirely agree 

 with either of the species described by Mr. Coquillett, but proved 

 to be the same as those observed in the act of ovipositing in the 

 buds. Moreover, from larvae inhabiting the buds at this time I 

 reared adults which, ovipositing in other buds on plants pro- 

 vided them, gave me eggs like those previously observed, as 

 well as larvae, pupae, and adults, afterwards reared from these 

 eggs. Clearly, I was dealing with a species whose larvae inhab- 

 ited rosebuds in the same manner as those described from 

 Washington. Specimens have since been submitted to the noted 

 British dipterologist, Mr. F. V. Theobald, who reported upon 

 them as follows: 



"I have made a number of careful preparations of the rose 

 cecidomyid. The female you sent previously had the antennae 

 broken. But for the antennae, they exactly answer to:£)ecido- 

 myia rosarum Hardy. I feel quite sure they are all Coquil- 

 lett's Neocerata rhodophaga (Bull. 22, N. S., Div. Ent., U.S.Dept. 

 Agr., p. 47. 1900). The males are all 9-jointed in regard to the 

 antenna?, and exactly answer Coquillett's description. I mounted 

 some twenty females, and found the antennal joints vary from 



9 to 11, so the character of his genus (which is certainly a good 

 one) in regard to antennal joints must be modified. There is 

 no doubt that antennal joints vary in cecidomyids, especially in 

 the female sex. In one I found an evidence of transition between 



10 and 1 1 joints. All the males, some fifteen, had the same geni- 

 talia and 9-jointed antennae. The globular second joint is very 

 characteristic. Note also the marked lepidopterous scale in 

 this species." 



While the normal antenna of the female is shown in Plate 

 III, i, and is certainly 10-jointed, with an occasional indication 

 of 11 joints, as stated by Mr. Theobald, the terminal joint being 

 prolonged and constricted but without distinct articulation, yet 

 there is a wider variation than was observed by him. In oneca se 

 a female was found with a 6-jointed antenna on one side, joints 3 

 and 4 being fused, with the terminal joint fully twice the ordi- 



