162 Knapiveed Gall-fly 



similar conditions. Of these seeds only fifty-seven germinated, equiva- 

 lent to 28-5 per cent. 



The percentage of germination in the seeds tested from galled 

 flower-heads was thus reduced 60-5 per cent, owing to the presence 

 of JJrophora larvae in the receptacles of the flower-heads ; probably 

 the nourishment that normally goes to the developing seeds, becomes 

 diverted to the growing larvae and also to the formation of the thick- 

 walled woody gall. 



This experimental result coincides with the indications given by 

 the appearance of the seeds usually obtained from galled flower-heads ; 

 many of these seeds are shrivelled, and smaller than the majority of 

 those collected from ungalled heads. 



Thus in addition to a reduction in the number of seeds produced 

 in the galled flower-heads, the number of seeds capable of germination 

 is also reduced. Taking into consideration the smaller number of 

 seeds obtained from infected flower-heads, and also the reduced vitality 

 of those actually ripened, as indicated by the above experiments, it 

 may be stated that the presence of the larvae of UrojjJiora solstitialis 

 in the flower-heads of Centaurea 7iigra reduces the number of eft'ective 

 seeds in these infected heads at least fifty per cent. 



Although the Trypaneidae or fruit-flies are justly regarded as 

 injurious from the economic standpoint because many of them are 

 very destructive to various cultivated fruits and to vegetable life in 

 general, yet those species whose larvae feed on, or in, the stems, leaves, 

 flower-heads and seeds of the weeds of cultivated land, may be regarded, 

 from the agricultural point of view, as beneficial. The result of the 

 activities of their larvae is that the plants fed upon are aflected adversely, 

 and in many cases, as in the case of the species considered in this paper, 

 the production of seeds is directly checked ; the distribution of the 

 weeds attacked is therefore restricted, even if only to a moderate extent. 



The question occurs as to whether destructive weeds could be 

 appreciably reduced by encouraging certain species of insects which 

 attack the seeds thereof. It is hoped that the present paper will suggest 

 to others the possibilities for further research and experiment in this 

 field of enquiry, which is fully worthy of attention from the economic 

 entomologist. 



