CLASSIFICATION OF GRACILARIA AND ALLIED GENERA. 161 



systematist attributes a little more or a little less value to some 

 slight variations of nervine arrangement. 



We may accept the result so far as showJhg a strong proba- 

 bility that the Gracilariad and Lyonetiad families are somewhat 

 related, and that the genera Opostega and Bucculatrix are more 

 nearly related to them than perhaps any others of the European 

 fauna. 



Meyrick does not divide them into families, and the series of 

 genera including Lyonetia and Gracilaria reads continuously. 



Interpreting this by his phylogenetic table (p. 708) by placing 

 gaps in the list of genera, it is a little less unnatural than it 

 looks; but when we go a little further, and find the connecting 

 link between Lithocolletis and Phyllocnistis is the terribly impos- 

 sible one of Argyresthia, we feel sure that that table has led 

 Mr. Meyrick into one of the pitfalls such tables open in all 

 directions, even for the most wary. 



Mr. Meyrick, however, discussed all these genera rather fully 

 in the ' Transactions ' of the Sydney Linnean Society in 1881, 

 with results that are but obscurely shown in his Handbook. He 

 recognizes, from geographical reasons, that Oracillaria is an 

 older form than Lithocolletis, and that the two groups are closely 

 allied from their larvae having the same number of ventral pro- 

 legs, viz. 6, in their post-Gracilarian stage. He does not, how- 

 ever, mention the Gracilarian trophi. I disagree with him when 

 he places Phyllocnistis with Cemiostoma, as he does still, and in 

 separating Bedellia from Lyonetia, and associating it with other 

 forms that are outside this group. 



It is rather outside this paper, but it may be noted that he 

 then placed Nepticida high in this group, and in the ' Handbook ' 

 he places it at the top of the group containing our Gracilarians ; 

 whilst the wing-structure and the pupal development both show 

 that it is as low as, or rather lower than, the lowest Adelidcs 

 (Tineae Aculeatae), though on a different line. Apart from these 

 points of disagreement, and taking into account that his classi- 

 fication is based almost entirely on imaginal characters, one 

 admires the grasp he shows of the generic relationships, and how 

 closely he approaches the true phylogeny. It must be remem- 

 bered that, in taking my own results to be more correct than his, 

 I am accepting his facts and conclusions to a great extent, whilst 

 modifying them by other series of facts ; and that but for the 

 sound and masterly foundation laid by Meyrick, Chambers, 

 and Stainton, my own contribution would have been of little 

 effect. 



Spuler and Rebel divide them into families. I am not in 

 possession of Lord Walsingham's classification of these genera, 

 beyond what I have been able to gather from a paper in the 

 'Proceedings' of the Zoological Society for 1897. From this it 

 appears that he associates together Bedellia, Bucculatrix, and 



