Xlll 



As this elaborate work is not likely to fall under the notice 

 of many of our readers, and Dr. Packard's detailed obser- 

 vations are too long to epitomise, we here content ourselves 

 with copying his sketch of the supposed order of the evolution 

 of the families of Lepidoptera from forms most nearly allied to 

 the Trichoptera, 



I have only been able to give a very brief sketch of some of 

 the various systems of classification of Lepidoptera which have 

 been proposed by various authors, without attempting to 

 analyse their work, or to go into details. Many more might 

 have been included, but brief and imperfect as it is, 1 hope it 

 will be found useful and instructive to those who may wish to 

 test any of the systems for themselves along the many paths of 

 inquiry which are open. For my own part I am rather inclined 

 to hope for a more natural arrangement of Lepidoptera by the 

 gradual progress of observation of their various characters and 

 stages, and the consequent slow re-arrangement of genera and 

 families, than by attempts to re-arrange the whole sequence of 

 families afresh, for this course can hardly fail to be the result of 

 laying undue stress on certain characters to the exclusion of 

 others equally important. This is one reason why authors who 

 have attempted this method have arrived at such discordant 

 results. Another is that every species has affinities not in two 

 directions only, but in many, and consequently any attempts 

 to arrange species in a linear series, which must be done in 

 a book or catalogue, really conceal almost as many affinities as 

 they emphasize. When all characters have been thoroughly 

 studied and compared, we can, perhaps, strike a satisfactory 

 average; but in the meantime, in so difficult a group as 

 Lepidoptera^ we are only justified at present in saying that 

 many of the existing systems of classification are about as 

 natural (or as unsatisfactory) as any of the rival systems. 



