63] LEPIDOPTEROUS LARVAE — FRACKER 63 



in the chaetotaxy of the prothorax. On Pseudanaphora, as on Hepialus, 

 the setae rho and epsilon are as far apart as alpha and beta or nearly so. 

 In Aegeriidae, many Tortricidae and others, rho has migrated forward 

 and has become associated with epsilon below alpha and gamma. One 

 or the other of these two conditions is present in all the Micros. In 

 Macrolepidoptera, on the other hand, epsilon has retreated caudad while 

 rho retains its primitive position below beta and delta. This character, 

 so far as I know, has no exceptions in those forms which bear primary 

 setae only. 



A definite and clear distinction is also found in the Kappa group 

 in front of the prothoracic spiracle. In the Macros this group consists 

 of two setae, kappa and eta, while all Micros except Pyralidoidea and 

 Zygaenoidea have three, theta being added. 



The crochets also furnish useful characters, varying greatly in the 

 more generalized half of the order but being nearly always arranged 

 in a mesoseries or pseudocircle in the more specialized moths and butter- 

 flies. Again the Zygaenoidea must be excepted. 



A point to which Dyar called attention twenty years ago is the close 

 relation of kappa and eta ("iv" and "v") on the abdomen and this 

 comes close to covering the Microlepidoptera as here limited. But ap- 

 parently this was not the generalized condition at all. Nearly all Lepi- 

 doptera have descended from species which would now be included in 

 the families Hepialidae, Micropterygidae, Acrolophidae, Tineidae, and 

 Yponomeutidae, and yet all these families have kappa and eta separated, 

 much as in Noctuidae. 



The following seems to be the best explanation of the conflicting 

 characters mentioned above: 



The setae of the ancestor of all Non-aculeata were arranged very 

 much as in Tineidae and Acrolophidae. (Figs. 7, 8, 15, 16.) Kappa 

 on the prothorax was a trisetose group, epsilon was located below alpha 

 and gamma and not associated with rho, kappa and eta on the abdomen 

 were wholly distinct, and the crochets were in a complete circle, probably 

 multiserial. Gradually evolution of this arrangement resulted in three 

 main divisions, of which it is now impossible to tell which is the oldest. 



One of these divisions (Figs. 39-42) retained theta on the prothorax, 

 but, above the most generalized forms, kappa and eta became associated 

 on the abdomen and all sorts of modifications took place in the develop- 

 ment of crochets. In these forms epsilon keeps its original position while 

 rho is likely to be found in any location back of it. The members of 

 the Tineoid series possess this variable set of characters and are often 

 honored with the name of "true Micros" in contradistinction to the 

 pyralids and zygaenids. 



