164 Knapweed Gall-fly 



larvae of at least three species of lepidoptera have been recorded from 

 Centaur ea nigra, the evidence adduced by Mr Morley that U. solstitialis 

 is a host of Pimpla sagax is hardly conclusive. It seems more probable 

 that the single specimen obtained by him either emerged from a lepi- 

 dopterous host, or that its presence in the flower-heads was accidental, 

 as he indeed suggests. 



Five hymenopterous larvae, probably of a Chalcid, were obtained 

 from 259 larval chambers examined in October, 1913, equivalent to 

 two per cent. Each of these larvae occupied a chamber that had 

 previously contained a Urophora larva. In one case there was direct 

 evidence that the hymenopteron had fed on the dipterous larva, as the 

 parasitic larva was inside a portion of the cuticle of the Urophora larva 

 and had evidently been feeding on the latter. From material in hand 

 I hope to be able to breed out a further supply of these Chalcids, to 

 get them identified, and to obtain the percentages. 



Remarks 07i Gall-formation and on other inhabitants of the 

 flower-heads of Centaurea nigra. 



The process of gall-formation in the flower-head of the host plant 

 exhibits some very interesting features. To determine whether the 

 formation of this gall is induced mechanically by the feeding habits 

 of the larva, or whether the latter secretes an irritant fluid which causes 

 the abnormal growth, would require further observations. The tissues 

 of the receptacle near the ovule attacked commence to swell up shortly 

 after the larva begins to feed, and in a comparatively short time the 

 size and shape of the future hibernating chamber of the larva are marked 

 out in the gall, although the larva is far from its full size. In a longitu- 

 dinal section of a gall that contained a larva about four weeks old and 

 1-3 mm. in length, the outline of the future hard-walled chamber could 

 be clearly distinguished. A thin layer of woody tissue is developed 

 within the hypertrophied tissue that the larva is feeding on, and a thick 

 layer of soft nutritive tissues is left within this woody layer and the 

 space occupied by the larva (Fig. 5) ; this nutritive layer is connected 

 at the base of the chamber with the vascular system of the plant. 



During the period that elapses between this stage and the attainment 

 of the fully fed one, the larva gradually eats away this layer of soft 

 tissue, until finally when the growth of the larva is complete the layer 

 of soft tissue has been absorbed and the hard wall of the chamber 

 reached (Fig. G) ; meanwhih' the wall becomes more lignified and 



