CLASSIFICATION OF GRACILARIA AND ALLIED GENERA. 139 



have been a Gracilaria, except for the want of the special Graci- 

 larian trophi. On obtaining these, in its first stage, it became a 

 Gracilarian. On passing this speciahzed larval condition on to 

 the second instar also, it became a Gracilaria or Ornix. 



I have examined a number of species of Gracilarianse {Graci- 

 laria and Ornix), and find they all have two stages with 

 Gracilarian trophi ; and the same is the case with the two or 

 three species of Ornix I have been able to examine. 



Mr. Chambers only observed one Gracilarian stage in this 

 group, but there are unquestionably two. 



Even in the first instars the habits differ in different sections. 

 Gracilaria syringella, for instance, with some half-dozen eggs 

 laid in a row, has the same number of larvae marching forward 

 abreast for some distance before they form a blotch. Gracilaria 

 sivederella and Ornix avellanella make a narrow thread-like mine, 

 which they lace to and fro into the pattern of a small square 

 gridiron before throwing it into one square blotch, which they 

 leave for a leaf-rolling, or perhaps I ought to say, cone-making 

 existence. 



In the genus Coriscium, which has always been associated 

 with Gracilaria, I am familiar with hrongniardellum and cuculi- 

 'pennellum, but on sulphurellum I made no observations. It is 

 described as cone-making, and is therefore congeneric with 

 cuculipennellum. 



Cuculipennellum is a true Gracilaria in its early stages, 

 having two Gracilarian instars, and afterwards inhabiting cones. 

 Its mode of pupation is very special, as it makes a cone such as 

 it makes for feeding in ; inside this it slings its cocoon like a 

 hammock, a structure very similar to, but more robust than, that 

 of Lyonetia clerckella, and it cuts out a little circular exit-hole in 

 tbe wall of the cone precisely opposite the end of the cocoon. 



Brongniardelliim has a very different history : the larva 

 makes a very large mine on the upper surface of an oak-leaf, 

 and there are often two or three larvae in a mine, but this is 

 always the result of the coalescence of as many different mines, 

 that happened to be in one leaf. It never leaves this mine until 

 it does so to pupate, and so far does not differ from such a larva 

 as that of Gracilaria omissella. It does differ, however, in a very 

 fundamental point, viz. it has three instars of Gracilarian form. 



In this it differs from all the other Gracilarias {sensu stricto), 

 and is allied to the Lithocolletid division, to which it ought to be 

 transferred. 



Though never leaving the mine until it does so to pupate, it 

 differs from such Gracilarias as omissella that do the same, in a 

 respect that again unites it to Lithocolletis and separates it from 

 Gracilaria. In feeding in its Gracilarian stages, omissella makes 

 a very small mine like other Gracilarias, and afterwards, wheij 

 armed with ordinary trophi, it mines in ordinary fashion beyond 



