6 MEMOIRS OF THE NATIONAL ACADEMY OF SCIENCES. 



of about 45°, the dull gray coloring of the wings with the lichen-green and flesh color giving 

 the whole such a perfect appearance to a piece of rough bark that the deception is perfect." 

 (Riley in Packard's Forest Insects, p. 269.) 



On the other hand the white ground color of the species of Cerura, with their black lines and 

 spots, make them very conspicuous, and it I'emains to be seen whether these are not warning 

 colors and whether they may not secrete a malodoi'ous or bad-tasting fiuid to render them 

 distasteful to birds. However this may be, the secretion is not sufliciently pungent or enduring 

 to prevent these moths from being eaten by Anthrenus larvi\? and other museum pests. 



ON THE LARVAL COLORATION OF NOTODONTID.E, AND THE ORIGIN OF SPOTS FROM LINES OR 



STRIPES. 



The observations hitherto made on the coloration and markings of Icpidopterous larvae, with 

 especial reference to their origin, have been those of Weismann and of Poulton on sphingid 

 caterpillars, and of Poulton and of Schroeder on the markings of georaetrid larvw. 



In preparing our monograph of the Notodontidai we reserved for a later occasion a discussion 

 of the origin of the stripes and spots, blotches, ".nd discolorations which serve to protect most of 

 the larva^ of this group from oliservation. What we have to ofl'er, however, ai'e only some facts 

 and conclusions derived from an examination of some living larva; and of the colored sketches 

 illustrating our work. 



Perhaps the mosi striking examples of the effects of lights and shadows in altering the color of 

 the green pigment of certain parts of the body is seen in the markings of smooth-bodied larvw which 

 feed among pine needles. In the larvoe of Nematus and Lophyrus, the green lai'vaj of geometrid 

 moths, also the larva; of certain Noctuid:e, including the Euro])ean Panoliv 2'>iniperda, also the 

 species of the sphingid genus Lapara (Ellenia), which habitually feed on the leaves of coniferous 

 trees, the white and red longitudinal stripes have evidenth' been produced by reflections from 

 the light and shaded portions of the leaf, while the red stripes are reflected from the red sheaths 

 of the needles. 



Humps a.s means of ohl'iterat'um or concealment. — Mr. Abbott H. Thayer truly says: "As 

 soon as patterns begin, obliteration of the wearer 1)egins." Thus a geometrid larva holding 

 itself out stiff like one of the twigs of its food plant, becomes lost to view. Some larva;, be 

 .says, "appear to be extensions of leaves," and this is admirably illustrated bj' the remarkable 

 caterpillar of JVerice hiclenfata (Monogr. Boinbycine Moths, I, Pis. XIX and XXIII). The pale 

 green and white shades of the bod}', the alternating oblique bars and stripes, blend with those of 

 the elm leaf on which it rests or feeds. This singular larva differs from all other known 

 notodontian caterpillars, and in fact from those of any other family, in each abdominal segment 

 from the tirst to the ninth, bearing a large fleshy two-toothed hump, the three largest on 

 segments three to Ave. Thus, as we have stated, "the outline of the back is serrate, and 

 perhaps mimics the serrate edge of the leaf of the elm on which it feeds;" the serrations or 

 humps are not only of the size and shape of those of the elm leaf, but the jagged outline of each 

 double hump strikingly resembles that of a serration of the leaf, which is two or three toothed. 

 Besides this, the tij)s of the teeth of the humps are reddish In-own, like those of the tips of the 

 leaf serrations. Moreover, the oblique dark and light lines leading up to the humps, as shown in 

 Miss Soule's excellent drawing (PI. XIX, fig. 4), strikingly resemble the light lateral veins of 

 the leaf and the shadows the^^ cast. 



With little or no doubt the tubercles and humps of the notodontians, of certain tree-inhabiting 

 noctuid larva% as well as many butterfly larvse, besides the geometers, are obliteration marks, 

 and have arisen in a M'ay difficult to explain, but evidently through a merely mechanical process, 

 and have been the means of giving a hold on life that unarmed caterpillars do not possess. 



The humped caterpillars of the notodontian subfamilv Heterocampinne, especially of the 

 genera Hyparpax and Schizura, are armed with high, often nodding, tubercles on the first, fifth, 

 and eiglith, or fi!-st and eighth, abdominal segments. 



While the nutant or movable tubercles evidently so function as to frighten away other 

 insects and thus ward oft' the attacks of tachina flies and ichneumons, these and the other humps 



