28 
MEMOIKS OF THE NATIO^s^AL ACADEMY OF SCIENCES. 
Neuroptera, Coleoi^tera, Mecoptera, Triclioptera, Lepidoptera, Diptera, and llymeuopteraj -where 
there are so many and pei'plexiiig eases of incongriieiice or divergence in larval forms whose 
parents are very closely allied. 
It is worthy of notice that in respect to Diptera the veteran dipterologist, Baron B. von Osten 
Sackeii, remarks of the nemocerous hies: “An arrangement of the imagoes based upon such prin- 
ciples will of necessity he justified by a more or less tangible correspondence in the characters of 
their larvm. This structural correspondence, this ijarallelism of larvm and imagoes among the 
Nemoccra^ suffers, as far as I know, but one exception, Mycetobia palUpes and Ehyphus. In both 
almost identical larvre produce flies belonging to different families.” (Berliner eiitomolog. Zeit- 
schrift, Bd. xxxvii, 1892, Heft iv, p, 418.) In the coi)y kindly sent me by the author a second 
case of Anopheles and Dixa is mentioned In the printed copy, but struck out by the author in the 
emended copy. 
Everyone is fiimiliar with the fact that there is a nearly similar incongruity between the larvm 
of the Muscidm and the flies. Many new facts bearing on this subject ai)p eared in Portchiusky’s- 
article on the habits of the necrophagous and coproi)hagous larvm of Muscidm, of which an English 
abstract by Baron E. xon Osten Sacken appeared in the Berliner ent. Zeitschrift for 1887. After 
speaking of the wonderful power of adaptation of these larvm to their environment, he states: 
Distinctly* related species belonging to different genera issue from larvic almost indistiuguisliablo from each 
other. And again closely related nnd almost indistinguishable imagoes, species of the same genus, differ in their 
oviposition (size and numher of eggs), and their larvjo follow a different law of dcveloinucnt (us to the degree of 
maturity the larva reaches within the body of the mother and the number of stages of development it passes tlirough). 
In one case even {Miisca corvina) larvie of the same species wore found to have a different mode of development 
in northern and southern regions of Kussia. 
Here also ifc is evident that the cause of the incongruity is due to the fact that the Jarvm, for 
the time being different animals from the adult, are modified by their euvlronmeut, the similar 
surroundings and habits of the larvm of (piite different genera causing the larvm e:^ternally at 
least to closely resemble each other. Whether they are so similar in their internal organs remains 
to be seen. Dr. C. W. Stiles, who has studied so carefully by microscopic sections tapeworms of 
externally similar form, and AAdiich cau not be separated by external characters, tells me that the 
internal organs seem to afford exeelleut st)ecific and generic characters. 
Lepidopterists in general do not hesitate to base their systems of classification on the larval 
as well as adult features. They in general regard their systematic arraugemeuts of the imagines fis 
more or less provisional, and all acknowledge that it is immensely satisfactory, oven after they are 
pretty well satisfied with th eir arrangement of the adults of a groiip, Avhether a genus or family, 
to work out the larval stages and to cheek their classifications based on adult features by the 
larval characters. Iii many cases they may be led to change the position of a species or genus, 
or to split up a genus or species. 
But, after all this, the fact that so many larvic, even in the same groiip, are hatched with such 
different shapes and characters; the fact that some are so much more simple aud primitive than 
others, opens up most perplexing yet interesting (xuestioiis and problems. We may, however, be 
able to solve these, and in the present groui> of Bombyces it seems to us that the different larval 
forms, some iirimitive and generalized aud others more oi' less modified or specialized, give clues 
to the phylogeuy of the groups which we confess we had not exi)ected. 
And in this memoir we have endeavored, though often it is mere guesswork, to drop the old- 
time method of putting the type species first and then ranging the others after it in an ill-assorted 
group, and have attempted to begin with what has seemed to us to be the ancestral form of the 
group, following with the later forms. This cau be best accomplished by taking into consideration 
the caterpillar, beginning Avith the generalized forms and ending with the later more modified 
or specialized forms. In such a large genus as Heterocampa this is not difficult to do. For 
example, as we shall see hereafter, the larva of IT. mantco is as simide aud generalized as any, 
while that of IT, unicolor is the most modified, with its semi-stemai>oda, firoui which Maerurocampa, 
with its fully formed stemapoda, may have descended. Aud then, while Cerura, Avith its stemapoda 
alike in all the species, is often or generally placed first in the group, it is evident that it was 
descended from some Heterocampa-like form through Macnirocainpa. Aided by our knowledge of 
