MEMOIRS OF THE NATIONAL ACADEMY OF SCIENCES. 279 



RECAPITULATION OF THE MORE STRIKING FEATIRES IN THE ONTOGENY OF CERURA. 



Conf/en it a I ch aractem. 



(1) The liirva batclies with fully developed steinapoda, indicatius' that the gemis has 

 'descended with little modification from a form like Macrniocampa. 



(2) The prothoracic horns are longer, better develoiied than in the mature worm, showing 

 that in this respect also the genus has originated from the Heterocaiiii)in:i'. 



Acquired characiers. 



(3) The head is smaller in ])roportion to the body than usual, owing to the great width of the 

 prothoracic segment. 



(4) The body is all brown above in the first stage, beginning to turn green in the second, and 

 in the third becoming nearly as in the last stage. Thus the colors are more diversified, with more 

 green in the fourth and fifth stages, rendering the now more e.x]>osed larva more adai)ted for 

 protection by the resemblance of its markings to the yellow and red spots on the green leaves of 

 its food plant, which appear early in autumn. 



(5) The dorsal hum]) on the third thoracic segment does not seem to appear until the last 

 stage. (Dr. Dyar, however, tells me that it appears in G. multiscriptd iu stage III.) 



(6) The filaniental legs retain their shape from the first to the last stage, but if anything are 

 •a little shorter in the last. On the other hand, the spinules in the third stage become larger on 

 the underside than before, the filaments being held curved up more than before, so that the 

 defensive spines on the underside, in response to external stimuli, have developed more ra]>idly 

 than those on the upper side. 



(7) Novel structures are the very long and well-developed supraanal plate and the pair of 

 •coproliferous spines (or dungforks) arising from the paranal lobes, and available for tossing away 

 the pellets of excremeut. These seem to be peculiar to the genus in this family. 



