328 IXSECT ARCHITECTURE. 



confessedly conjectural ; for tliougli Swammerdam detected 

 a gall-fly in the act of depositing her eggs, he did not attend 

 to this circumstance ; and in the instances which we have 

 observed, some unlucky accident always prevented us from 

 following up our observations. The indefatigable Eeaumur, 

 on one occasion, thought he would make sure of tracing the 

 steps of the process in the case of the gall-fly which pro- 

 duces 'the substance called hedeguar on the wild rose-tree, 

 and to which we shall presently advert. His plan was to 

 enclose in a box, in which a brood of flies had just been 

 produced from a bedeguar, a living branch from a wild 

 rose-tree ; but, to his great disappointment, no eggs were 

 laid, and no bedeguar formed. Upon further investigation, 

 he discovered that the brood of flies produced from the 

 bedeguar were not the genuine bedeguar insects at all, but 

 one of the parasite ichneumons ( Callimone Bedegiiaris, Ste- 

 phens), which had surreptitiously deposited their eggs there, 

 in order to supply their young with the bedeguar grubs, all 

 of which they appeared to have devoured. It may prove 

 interesting to look into the remarkable structure of the 

 bedeguar itself, which is very different from the globular 

 galls above described. 



One of the bristles of the Bedeguar of the rose magnified. 



The gall-fly of the willow (^Cynips viminalis) deposits, as 

 we have just seen, only a single egg on one spot ; but the 

 bedeguar insect lays a large cluster of eggs on the extremity 

 of a growing branch of the wild rose-tree, making, probably, 

 a proportionate number of punctures to procure materials 

 for the future habitation of her yoinig progeny. As in the 

 former case, also, each of these eggs becomes (as we may 

 suppose) surrounded with the sap of the rose, enclosed in a 

 pellicle of gluten. The gluten, however, of the bedeguar 

 insect is not, it would appear, sufficiently tenacious to con- 



