246 THE CANADIAN ENTOMOLOGIST. 



group terms, and many so-called families are absolutely incapable of 

 definition. A great many questions come up for settlement in the pre- 

 paration of this list of Lepidoptera. Besides the editing, my work has 

 been confined to the macro-heterocera as far as the Geometrina, and on 

 the families there embraced a few remarks are made as explanatory of my 

 views on the subject. 



The Sphingidce are well limited, and the classification is probably on 

 a tolerably sound basis, yet there are two very distinct series that are 

 probably independent branches from the original Bombycid stock. The 

 Smerijithince now have their nearest allies in the Ceratocampidce, while 

 the Sphmgiiue have their allies with the Notodontidce. The Sesiidm, 

 which custom has placed immediately after the Sp/iingidce, have nothing 

 whatever to do with them. They are different in all stages, and save for 

 a superficial resemblance in the imagines, there is no relation whatever. 

 They should be associated with the Cossidce, from which they are to-day« 

 very distinctly separated ; yet, besides the larval habit, the imagos in 

 many of the Sesiid genera have real Cossid affinities. 



The Thyridee are few in number, and the family is well marked. 



Under the term Zygcenidce a most heterogenous mass has been usually 

 classed. I long ago pointed out that none of our American species had 

 the structural characters required by the definition of the family, and that 

 the family, so well represented in Europe, is totally unknown in boreal 

 America. I have, therefore, split up the series into compact groups, 

 with family appellations. 



The Agristidce perhaps do not deserve to be called compact, and 

 possibly part of the series may yet be transferred to the Noctuidce. 



The Syntomidce are very well limited, but only feebly represented 

 with us ] nor have we many of the more typical forms, most of them 

 resembling in appearance more nearly the PyromorpJiidce. The latter 

 family is also small, well defined, and nearest to the next, the Ctenuchidce, 

 also a well limited group. The Fericopidce form a natural lead into the 

 Arctiidm and Lithosiidce ; and, indeed, beginning with the Syntomidce 

 and running through the Arciiida, we have a fairly natural group. The 

 Fericopidce, are known to me only in the American species, and the 

 family reference is by Mr. Hy. Edwards. 



Three genera, which are probably related, are placed in the Hetero- 

 gynidce. This is also done on Mr. Edwards's authority, and I do not 



