26 



C. Ewpis (Empi?mc). Small species as yet undetermined. 

 Prey nearly always minute Diptera, chiefly Cecidomyia and 

 Psychodcs. 1909 — two; 1910 — fifty-five; 19 11 — one hundred 

 and three. 



D. Ewpis tessellata. Prey very varied, but always Dip- 

 terous. 1908 — two; 1909 — two hundred and twenty-four; 

 1910 — twelve ; 191 1 — thirty-three. 



E. Empis opaca. Prey like that of tcsscllata, but mainly 

 of the genus Bibio. 1909 — one hundred and sixty-eight; 

 1910 — forty-six; 1911 — fort)'. 



F. Empis livida. Prey more varied than that of any other 

 species of the genus, but still chiefly Dipterous. 1908 — three 

 hundred and five, of which 4 were collected by Mr. C. II. 

 Ilamm ; 1909 — forty-five; 191 1 — thirty-two. 



III. The prey or object provided by the male is not 

 devoured by the female, but becomes as it were an ornament 

 or plaything providing some indispensable stimulus. 



A. Hilara (Empiuac). Many species as yet undetermined. 

 All the species fly over water, and the prey or other object is 

 always picked up from its surface by the male Hilara. The 

 males take floating insects of all kinds — sometimes specially 

 Diptera, sometimes Aphids— scales off overhanging trees or 

 other fragments of plants. Some of the species will accept 

 almost any floating object, while others seem to restrict them- 

 selves to particular insects, Mich as Apliidac. When the 

 object is very heavy the male, after seizing it, spins round 

 with great velocity till the load rises on a cone of water and 

 is finally lifted from the apex. In Mr. Ilamm's experiments 

 disabled Diptera of the genus Chironcmns, &c, stamens of 

 buttercups, and ray florets of daisies strewn on the water 

 were soon taken by the males and afterwards found in the 

 possession of the females. Pairing invariably occurs upon the 

 wing, but numbers of specimens show that a sweep of the net 

 through the swarm at first catches nothing but males carrying 

 the objects that had been strewn on the water, while a later 

 sweep catches pairs still carrying the same objects. The 

 specimens illustrating this investigation are all carefully 



