22 ILLINOIS BIOLOGICAL MONOGRAPHS [22 



Series of changes. Ordinarily two setae in the same relative position 

 on their respective segments are considered homologous. Two setae in 

 different positions may be homologous, however, if they have reached 

 these new positions by migrations during phylogeny. In many cases 

 an intermediate location is shown by the first instar, while in others a 

 large number of mature larvae will show a complete series of steps in 

 the migration from the old position to the new one. 



CHOICE OF A NOMENCLATURE 



At the present time Roman numerals are in more or less general use 

 as a means of denominating the setae of lepidopterous larvae. They 

 were introduced by Dyar and have been retained, largely in their original 

 sense, partly because the great majority of recent papers describing 

 caterpillars have been written by this worker. The wide distribution 

 of his papers and the undoubted accuracy of his observation would lead 

 the present writer to adopt his nomenclature if that were possible. For 

 several reasons, however, it seems that the use of numbers in this con- 

 nection is no longer desirable. 



In the first place the abdomen of Frenatae, upon which the numeral 

 system was based, can not represent the original type. In some particu- 

 lars it differs radically from the same segments of the Jugatae. Attempts 

 to derive the setal plan of the thoracic segments from this arrangement 

 and to find any evidence for such a derivation have been fruitless. The 

 original type segment, or better, the potential arrangement from which 

 the modern plan has been derived, must have included a greater instead 

 of a smaller number of setae than are at present found on the meso- and 

 metathorax. Many of the thoracic setae are undoubtedly as primary 

 and as ancestral as those on the abdomen, and the same is true of some 

 of the additional setae shown on Hepialus; therefore, if Dyar's numerals 

 are used we should be compelled either to give letters and subnumerals 

 to primary setae simply because they are not present on the abdomen 

 of modern caterpillars, or to adopt his scheme for the abdomen of Fren- 

 atae and apply a different one to all other body-segments and to the 

 Jugatae. Neither of these methods would fulfill our hope of a uniform 

 nomenclature based on homology and applicable to all the segments of 

 all lepidopterous larvae. 



In the second place any series of names which have as definite an 

 order as numbers is sure to be misleading and is likely to prejudice one's 

 views in regard to homology. The mere use of numerals beginning at 

 the mediodorsal line tends to cause one to give the upper of two setae the 

 smaller numeral and to neglect the fact that some setae are absent. In 

 wing venation, it is found that when numbers are used, workers tend to 



