OP WASHINGTON, VOLUME XII, 1910. 63 



A little farther on I found both species again upon Solidago 

 canadensis L., and Geranomyia rostrata, which all through 

 the glade outnumbered the other species, also upon Solidago 

 rugosa Mull. 



A week later, in an excursion to Lakeland, Maryland, I 

 observed Geranom\ia again; this time only G. rostrata. They 

 did not appear until near sunset, and this later appearance here 

 I attribute to the open character of the region. They were 

 again upon the flowers of Eupatorium pnrpurcnni and in ad- 

 dition they visited the flowers of Helianthus strnmosus L. 

 Still a week later, on October 3, I found an individual of Ge- 

 ranomyia canadensis upon a flower of Aster cordijolins I_. in 

 Rock Creek Park (Washington). 



Two further records are due to the kindness of Mr. W. L. 

 McAtee. He observed Geranomyia canadensis on the flowers 

 of Eupatorium ageratoides L. at Plummers Island, Maryland, 

 September 12, 1909. Geranomyia dirersa O. S., apparently 

 a rare species, was taken by him at Beltsville, Maryland, Au- 

 gust 4, 1909, upon the flowers of Solidago canadensis L. 



It is worth noting that all these visits of Geranomyia, those 

 recorded by others as well as those observed by myself, were 

 to flowers of Compositae. The flowers of this group of plants 

 are well understood to be so organized that they are profitable 

 only to insects which are specialized to reach the nectar at the 

 bottom of the deep and narrow tube, this latter protecting the 

 honey from undesirable marauders and rain. When we con- 

 sider, then, the character of the flowers visited by Gerano- 

 myia, its proboscis appears clearly as an adaption to flowers, 

 moreover to those flowers which offer a superior quality of 

 honey to those insects organized to obtain it. It is rather sur- 

 prising, with these facts before us, to find that Graenicher, in 

 his study of the flower-visitors of Composita?, classes Gerano- 

 mvia among the "allotropous" visitors; that is, those which 

 are occasional, show little or no adaption to the flower, and 

 are unimportant to the plant. As to this last I cannot answer, 

 but certainly Geranomyia, with its long proboscis, is special- 

 ized to probe into flowers with deep and narrow honey-tubes. 

 Quite likely the close relation of Geranomyia to flowers has 

 escaped notice on account of its crepuscular habits. It was 

 only after I became aware of this that I was able to make 

 more conclusive observations. 



A number of Tipulidce of other genera have been recorded 

 as flower-visitors by students of the fertilization of flowers by 

 insects. Knuth, who has brought together the records of in- 



