edge of the amount and kind of variation in structure in the 

 group. 



There is, finally, something real in family form, as elucidated 

 by Agassiz, which Dr. Packard, in the moths, finds in head- 

 characters and Dr. Herrich-Schaeffer in wing-characters. Lederer 

 seems to find an average from all parts of the ringed butterfly or 

 moth ; his " families " differing more from the combination of 

 characters they present. Exceptions to any one family char- 

 acter, such as disproportionate size in the whole insect or any of 

 its parts, do not appear to me to invalidate the conception of a 

 family. I do not wish to enter into argument as to the best 

 classification of the TineidcB, but, disagreeing with Mr. Chambers, 

 I do not think any one would take Anaphora for anything but a 

 Tineid. There are difficult genera on the confines of almost every 

 family, but I think that alter a complete knowledge of all the 

 characters, and the neuration is a very important one, we shall 

 finally be able to agree upon the best position for any moth we 

 have discovered yet, and this without establishing any family to 

 rank with the Pjralidce, Tortricidce or Tincidce (excl. Alucitae), 

 as now considered. So far as we know we are not warranted in 

 considering the Tineidce as consisting of an assemblage of de- 

 graded forms, but a natural group affording sub-families not dif- 

 fering more widely on the whole from each other, than do the 

 sub-families of Pjralidte, as arranged in my " New Check List." 

 When we come to the study of descent we arc without paheonto- 

 logical evidence, in fact, without any evidence which would 

 justify our considering the long cilise as degradational. We have 

 some evidence that a short tongue is a character of an old type 

 and that the long maxillae have come into existence with flower- 

 ing plants. Thus, that the Bonibyccs are a group of sub-families, 

 descended from a short-tongued early group of Lepidoptera, is a 

 reasonable supposition. Aquatic habit, more or less correlated 

 with structure, may be assumed as degradational, since it is ex- 

 ceptional in different families, and a water-life is lower than a 

 land-life, as we see that it obt.iins in embryonic conditions and in 

 low groups of annulose animals. In this connection Professor 

 Comstock's discoveries of aquatic and predaceous larvae in our 

 moths are of immense importance. Great interest attaches to 

 special enquiries of this kind, i. e., as to the value of a larger or 

 smaller number of abdominal fleshy feet to the larvae. So far as we 

 can judge of degradational characters from biological data we have 

 a solid basis for our deductions. The absence of testimony from 

 fossils renders a certain class of speculations valueless. To return 

 to the subject of classification, we cannot afford to dispense with 

 neurational characters; hence, I should object to Mr. Smith's 

 recent synopsis of the Noctiiidce, upon the ground that he dis- 

 tinctly passes over this important branch of the study. Although 

 it is uniform, to a large extent, the neuration of each genus should 



