MK.MOIKS OF TlIK NATIONAL ACADEMY OF SCIFNCES. 39 



wise nioditiod or specialized, iiiid ciiiivcrtiMl into an enormously lone' instfuinent for ])i'oliiiiu' the 

 deep tuliuiui' corollas of orchids. 



I'lie jialpi of Sj)hina'itla' ai'c always ihrcc-jointcd and rallicr larLi'c. while those of th(> C'erato- 

 campiiia' have lost the third joint, hut in the Auliina' and l>nna'in:e they are hirge and often 

 three-jointed. 



The legs of Sphiiijj'idie are stouter and piox ided with stout spines, tiuis ditiering from those 

 of the Ceratocampida'. wiiieh are comparatively weak and unarmi'd. 



The genitals of Sphingida' are in certain genera like those of Ceratocampidre and do not 

 present family distinetions. 



We have seen that there is in reality hut a slight break or gap between the Ceratocampidaj 

 and Sphingid;e. Were it not for the changed positions in the larva of tubercle i\ the presence 

 of an additional branch of the radius vein strengthening the costal edge of the fore wing, and 

 of the frenulum, there would be no absolute characters separating the primitive Sphingidro 

 (Smcrinthin;e, especiallj' Cressonia) from the Ceratocampidiv. The larger head, fusiform or 

 prismatic anteonfe, long maxillse, narrow, strong wings, stout, spiny legs, and the slight differ- 

 ences in the larva and pupa are simply adaptive characters, due to exercise of the modilied organs, 

 the result of the greater activity of the imago in seeking its food, in probing the deep corollas 

 of flowers, and in seeking their mates. 



The I'ei'h'al, restoratio)), or rcaetpdsition of pari) ally atrophied organs. — We have observed 

 what a great range in size and adaptability for probing tubulai" corollas is seen in the develop- 

 ment of the maxillae of the Sphingida?, from an almost rudimentary condition in Cressonia to 

 those of the common potato Sphinx fPhlegethontius), and to the enormously long one of the 

 South American Sphinx, and that this is evidently the result of u.se, and of use-inheritance. 

 This is correlated with the narrow, powerful wings, the large thorax due to the enlarged muscles 

 which raise and lower the wings; with the stout spiny legs, and the large head. 



If, as we have attempted to show, by presenting the facts supporting the view that the 

 family of Sphingidse has directh^ descended from some member of a definite family, i. e., the 

 Ceratocampida?, then we have to deal with instances of a most remarkable phenomenon, that of 

 the revival, restoration, or bringing back to active use, and consequent increased development, 

 of organs or structures which in the ancestral or stem forms have become partially or almost 

 wholly atrophied from disuse. It is universally the case that an organ, once wholly atrophied, 

 never becomes restored or revived so as to function or be of any service in the animal economy. 

 We have seen that in the case of the wings, a branch of the radius vein (III) either entireh' 

 atrophied or only vestigial in difterent groups of Saturnoidea becomes greatly developed in the 

 Sphingida?, thus strengthening the costal edge of the wing. This is a clear case of the restoration 

 oi- reaccjuisition by exercise of a structure or organ. 



Another case is that of the maxillse. We should regard those of Cressonia as rudimentary 

 rather than vestigial: but those of the Saturnoidea are, as anyone will acknowledge, vestiges of 

 organs which, in the ancestors of the group, were well developed and of constant use to the 

 insect, as in the Noctuidse. It follows from this that here we have an instance, and we know of 

 none others on record, of the complete revival or restoration of the muscular, nervous, and 

 mechanical power and activity of a lapsed or nearly atrophied organ. 



The infinite variety in the morphology of the mouth-parts of the arthropod phyla does not 

 afford, so far as we are aware, .such an instance. It is a nearly universal law that an organ in 

 the last stages of atrophj' is never restored to its pristine structural and functional activity. 



To suppose that hj any process in nature the lost digits of a horse could ever be restored, 

 and that the splint bones could in the descendants of the modern horse in future ages be restored 

 and function as usable toes, seems on the face of it an absurdity; and yet in the useless tongue of 

 the ceratocampid moths we have, unless we are mistaken, an organ which, in the descendants of 

 the group, has become restored in form, structure, and vigor, and so greatly enhanced in develop- 

 ment as to form a most striking case of restoration by simple exercise maintained through many 

 jrenerations. 



