[ 162 ] 



Great Britain. It is recorded from Devon, Hereford, Hants, Kent, Radnor, 

 Middlesex, Hampshire and Inverness-shire. We have found it in Northumber- 

 land and Aberdeenshire. 



The adults of this fly are recorded over the period from May till Sep- 

 tember. Schiner (1862) writes regarding the genus: " The flies live in dry places 

 especially heaths. They are also found on Umbelliferae and on waste ground 

 are often to be seen in large crowds on Dauciis carota." He also states that 

 B. geniculata visits Erica vulgaris. Walker (1853) reports these flies as "very 

 common." 



The following list of flowers known to be visited by B. geniculata is com- 

 piled from Knuth's Handbook of Flower Pollination (1909). 



Viola lutea. Valerianella olitoria. 



SteUaria holostea. Succisa pratensis. 



Medicago lupulina. Eupatoria cinnahinum. 



Potentilla sylvestris. Pulicaria dysenterica. 



Potentilla sterilis. Myosotis sylvatica. 



Daucus carota. Mentha aquatica. 



Hedera helix. Origanum vulgar e. 



Asperula cynanchica. Hottonia palustris. 

 Asperula odorata. 



METHODS OF INVESTIGATION. 



The Tipula larvae, after being washed free of soil, were examined individu- 

 ally and with a Uttle practice the parasitic maggot could be readily recognised 

 beneath the skin of the host larva as an elongated yellowish patch of definite 

 shape. 



The proportions found to be infected varied a good deal in different 

 localities and in different years. In one area, viz. Echt, Aberdeenshire, the 

 niunbers reached a significantly high figure. Over a period of one month, 

 during which larvae were examined, a progressive decrease in the percentage 

 of infected individuals was noted. This was found to be due to the fact that 

 as the season advanced, the infected larvae were dying off, and pupae of 

 Bucentes were becoming increasingly more numerous in the soil. It must 

 therefore be borne in mind that for the first generation, unless examinations 

 are made not later than about February, any estimate of the degree of infection 

 must be below that actually existing. The following data illustrate this : 



