MEMOIRS OF THE NATIONAL ACADEMY OF SCIENCES. 45 



As roo-jirils ilu' iiiiui;'!) --tutc, tlu' liciul, wlioii tlcmulcd. of notniloiit iaiis is iiiiicli us in Cei'ato- 

 cuinpiniv, tlio opicrauimn being trisuigular. (Compare I'l. XL\'III of this inoriograph, Pt. I., 

 with PI. XLIV.) In voiiation tiioro is a goticral roscmhlanco, though in the Cenitocampiii.T there 

 is a loss of two radial veins. Aside from any theory, the Notodontichv proper are closcdy allied to 

 the Ceratoeanipidie. dirt'ering- in characters which, on the whole, are more primitive. The rela 

 tions of the .Suturniida^ to the Ceratocampida^ will be tliscussed in a future part of this work; 

 ill the diagram (tig. -J-) the position assigned to tlie HemilencidiW is purely tentative and 

 ))rovisional. 



It may be objected that the Notodontidie, especially the Ileterocampinie, markedly diHer 

 from the Ceratocampida* in not possessing the large usually tuberculated suranal plate and the 

 very large anal legs or claspers of that group. It should, however, be observed that in several 

 species of Heterocampa the long triangular suranal plate either bears a pair of horns, or at least 

 three pairs of setiferous tubercles. On the other hand the anal legs of these notodontians are 

 much specialized in the direction carried out so remarkably in the Cerurinie. Hence we ai'e led 

 to suppose that the Ceratocamijinw sprang from a more generalized form, which on the one hand 

 gave origin to the Ceratocampida^ as a whole, and on the other to the existing species of Hetero- 

 campa, of Macrurocampa, and tinallv of Cerura and its allies (Cerurina^). 



Genera, fain il Us. etc., aiitfickil groups. — An}' systematist after finishing his work on a group, 

 rises from his task impressed with the difficult}- of classifying the genera and species. This 

 embarrassment is, of course, due to the fact that his best etl'orts are only tentative and provisional 

 attempts to trace out the intricate and bewildering lines or network of affiliated forms. Evolu- 

 tion has gone on by divergent paths, the lines of development forking and reforking from a 

 common origin. Without the aid of the theory of descent, without seeing that everywhere there 

 is a progressive development from the generalized or primitive to the specialized or more recent 

 forms, we should be lost in the fog, or be like sailors without a compass. If we believed that 

 variation was indefinite, fortuitous, without reference to changes in the conditions of life, we 

 should indeed be still sooner lost in a tangled mass of forms, especially in a great group like the 

 Lepidoptera. • 



To take the present case as an example. The Notodontidiv, as compared with the families 

 evolved from it, is a composite or synthetic group, the mother of at least nine families, while 

 from the highest or most specialized family, the Sphingidaj, or some unknown group or form 

 allied to it, the Castniidie and the dift'erent families of butterflies forming the great superfamily 

 of Papilionides have probably evolved. 



So far from being a homogeneous group or family, we have seen that within the limits of 

 what we call the Notodontid;¥ there are two chief groups, one with larva" armed with multisetose 

 warts, and one with larv;v provided with tubercles giving rise to a single seta or bristle. 



From the beginning of the world's history, for we see it in Cambrian fossils, not only a 

 tendency to. but also an actual and rapid or tachygenic process of modification in diti'erent direc- 

 tions has taken place. Just as soon apparently as what we call the Notodontid;e ai-ose from 

 pi'obably the small group of Thyatiridaj, it began to diverge, to spread out and become adapted 

 to different conditions. While the more normal forms liecame Xotodontida\ there arose, follow- 

 ing the lino of least resistance, as the result of adaptation to conditions not encountered by other 

 forms, the more aberrant genera Hyparpax, Heterocampa, and Macrurocampa, as well as the 

 Cerurinw. 



Meanwhile, in certain forms the tubercles became flat and broad, divided into a numl)er of 

 heads or tubercles, each bearing a hair, and gave rise to the genera Icythyura, Datana, Apate- 

 lodes. Datana Ijecame a specialized closed type, represented by numerous species, while 

 Ichthyura became the parent or stem-form of dve important and in some cases numerically 

 successful families; Apatelodes, .so far as we can judge, having given origin to the Euryptero- 

 tid;¥. If this phylogeny should prove incorrect, there is strong circumstantial e\idence that the 

 groups arose from similar though extinct forms. At all events, evolution followed various lines, 

 determined by the various conditions of life, and these lines are older or more recent, shorter or 

 longer, more or less divergent, each type adapted to its particular niciie. habitat, or mode of life. 



