290 [Deckmbek 



brood of forty, which I carried through to pupation, only one (the last 

 which matured, later by four days than any other) was of the ordinary 

 green, but without the usual change of shade, before burying in the ground. 

 This marked departure from the characteristic coloration, was doubtless 

 the result of indoor development, for I have before observed similar dii^er- 

 ences in color in larvae which I have reared, from those of the same spe- 

 cies matured with their natural exposure, — the variation tending always 

 to deeper shades, — the reverse of what would naturally be expected, rea- 

 soning from the known eflFect of sun-light and air upon the vegetable world. 



In many of the Sphingidae the larval coloration is subject to such vari- 

 ation, that it is of comparatively little value in the description of species. 

 For instance, in Sphinx quinquemaculata Haw., we find all the intermed- 

 iate varieties of color, from a grass-green to dark brown: S])lii%x cingulata 

 Fab., is said to present the same range of variation. The external larval 

 Btructure, however, is constant, and therefore of primary importance. 



In C. quadricornis, the head is large, semi-oval, flattened in front, gran- 

 ulated. Body with the abdominal segments cylindrical, thoracic segments 

 tapering slightly; on the second and third segments each, two short, straight, 

 tuberculated, fleshy horns; a dorsal row of fleshy serrations, commencing 

 on the fourth segment, of one tooth to each of the eight wrinkles,* in 

 which the segments are divided; from the anterior horn, a short row of 

 tubercles extends downwards to a faint subdorsal line, sometimes obsolete; 

 a line of granulations connects the thoracic horns, and is continued indis- 

 tinctly across the fourth segment; a similar substigmatal line crosses the 

 second and third segments, and joins the first lateral stripe; a line of tu- 

 bercles borders the collar anteriorly; seven lateral stripes of granulations 

 extending over the whole of one segment, and two-eights and six-eights 

 respectively of two others. Caudal horn of medium length, stout, curved 

 and quite rough with acute granulations; anal plate and shields granu- 

 lated. 



On the 9th, four of the larvae reached maturity, indicating the period 

 by ceasing to eat, and remaining motionless in one position. The follow- 

 ing day they descended from the branches and commenced crawling vig- 

 orously about the case in which they were confined, endeavoring to make 

 their escape. At this stage, the larva cannot be induced to enter the 

 ground, and even if covered by a body of earth, it forces its way to the 



■••■ Thi3 number is uniform in all the Sphinx larvae which I have examined. an<l 

 probably in all those which are wrinkled tran.sversely. 



