23] LEPIDOPTEROl'S LARVAE — FRACKER 23 



neglect studying out the true homology. This danger is still more appar- 

 ent in work on the setae, for whereas it is rare to find a branch of radius, 

 for example, crossing media and reaching the margin of the wing back 

 of the latter, it is an ordinary occurrence for seta ii to be found above 

 i, for seta v to take almost any position with respect to iv, and for vii 

 to consist of one, two, three, or many setae, either approximate to each 

 other, or decidedly remote in position. 



In place of the numerals now in use the writer suggests Greek letters. 

 They combine a quickly written character for labeling plates and an 

 easily pronounced word with which all scientific men are familiar. A 

 special letter can be introduced for a subprimary seta in a limited group 

 without disarranging the system. The alphabetical order is not so fixed 

 in the mind as to prejudice one in regard to homology. At the same 

 time the confusion introduced by the papers of Quail and Forbes in 

 Dyar's system is avoided and, as long as the Greek letters are retained 

 in the original sense, the meaning will be absolutely clear. 



As it is convenient to have a single term for groups of setae and for 

 compound tubercles, such as scoli, the writer capitalizes the letter indi- 

 cating one of the more constant of the primary setae composing the 

 group. For example a verruca bearing a tuft of setae developed from 

 the group consisting of theta, kappa, and eta, is known as the verruca 

 of the Kappa group and labeled "K" (ef. Figs. 7 and 25). In cases 

 where there is doubt about the number of primary setae from which a 

 scolus or verruca is developed, the capital letter is usually employed. 

 This method is particularly valuable where it is impossible or disad- 

 vantageous to indicate the components of the group. The names of the 

 groups and the setae composing them are given on page 39. 



As mentioned on a previous page, a seta is known as primary when 

 it is present on the newly hatched larva. If it first appears after one 

 of the molts but is fairly constant, it is known as subprimary, e.g., mu, 

 theta, etc. Secondary setae are those which have no constant position, 

 are more or less numerous and scattered, are not ancestral, and bear 

 absolutely no relation to the primary setae. They are very rarely found 

 in the first instar. The individual setae of a tuft borne on a verruca 

 are not given distinctive names, the group itself being called primary or 

 subprimary according to its origin. The numerous setae which it bears 

 are not considered secondaries. 



IV. Setal Arrangement in the Principal Superfamilies 



In this section the larval chaetotaxy of a number of typical species 

 is described. Each seta is named when it is first mentioned in the 

 description, the same name being applied to its homotypes on other seg- 



