l.«64.] 637 



itis allies brown with eye-like siibterminal spots; while Mrlltxa and Ar- 

 <///)inis are fulvous or fulvous red above, with crenulate lines and lunules 

 of black on certain fixed parts of the wing ? 



Again, it is difficult to conceive of any peculiarity in structural or- 

 ganization, which can account for the wonderful phenomena of Phy- 

 TOPHAGIC Unity ; why, for example, Cynips should form galls on the 

 Oak and never on the Rose, and Rhodltes should form galls on the Rose 

 and never on the Oak; why Pontia and Pirn's should aifect crucife- 

 rous plants, CoUas the clovers, Parnassius the saxifrages, and Argynnia 

 the violets. We find that, even within the boundaries of the United 

 States, the gall-making genus Gecidomyia inhabits at least 8 distinct 

 genera of plants. (See above p. 552.) Why are the gall-making genera 

 Ct/nips and Rhodites each restricted to a single genus of plants ? We 

 find that At-ctia and its allies are very generally polyphagous, and feed 

 on an almost unlimited number of different genera of plants. Why is 

 Arctitt polyphagous, and Pontia and /*/firw and Co^ms and Parnassius 

 and Argynnis generally monophagous? It is inconceivable to me, that 

 in genera all belonging to the same Order, as with these last, there can 

 be fundamental and immutable diiFerences in the structure of their 

 mouths or their stomachs, of such a nature as to enable the one to eat 

 and digest almost anything of a vegetable nature, and to compel the 

 others to restrict themselves, as a general rule, for thousand and thou- 

 sands of years to one single g'fenus of plants. Look at the exclusively 

 American Lepidopterous family Dryncampadse. Within the limits of 

 the United States there are now known to be eight, or in any case seven 

 species belonging to this family. Six (or five) of them belong to the 

 genus Dryocampa. and out of the six (or five) no less th-An fo)ir , pelhu^i- 

 da, senatoria, stigma and bicolor — or, at all events, if hicolor be not, as I 

 believe it to be, a true species, no less than thn-e — inhabit the Oak 

 in the larva state. Of the remaining two, ru/jicuu.da, which in- 

 habits the Maple, is rather an aberrant form, and imperialism which 

 inhabits the Sycamore (Platanus). the Pine, the Sweet-gum (Liqui- 

 damber) and occasionally the Oak, is a decidedly aberrant form. The 

 other two genera of this family, Gcratocanipa Harris and Sphingicampa 

 Walsh, each containing one N. A. species, are, but more especially the 

 latter, pre-eminently aberrant forms ; and it is most remarkable that 

 neither of them has ever been found on the Oak, the former feeding 



