18 MEMOIRS NATIONAL ACADEMY OF SCIENCES. vol. xii, 



Larva. — Stage I: For our knowledge of this stage we are indebted to Mr. Poulton's 

 careful description and excellent figures. 



Length, 6.25 mm.; at the end of the stage, 10-11 mm. Head green and rounded. The 

 ground color dark green; along each side a white subspiracular line ending in front on the 

 base of the prothoracic spine, and behind ending at tbe base of the red terminal suranal spine. 

 There are seven complete oblique lateral white stripes sloping in a posterior, a direction the 

 reverse of that of most Sphingida;, though in Sesia there are similar reversed stripes. The two 

 prothoracic dorsal are slightly shorter than the metathoracic dorsal horns, but are much as in 

 stage III and all a little longer, but otherwise like the caudal horn; all are bright red. He 

 describes the head as very generalized, and like that of the young form of Smerinthus and 

 Sphinx. The length of the stage is unusually long, i. e., 16 days. 



Stage II: Length 10.5 mm.; at end of the stage 15.5-17 mm.; color and marking almost 

 identical with those of the previous stage, and the terminal fork on the five spines not so marked 

 in this stage, and it becomes less so in the next, the fork being entirely absent in the last stage 

 and often in stage II. Length of stage about nine days. 



Stage III: Length 16 mm.; at end of the stage 22.25-27 mm. Differs but slightly in color 

 or armature from stage I. "The anal red spine is relatively shorter and less conspicuous" 

 (Poulton). A blown larva 23 mm. in length, and evidently in stage III, I have received from 

 Staudinger-Bang-Haas. 



Prof. Poulton does not enter into details regarding the armature. Notable features are the 

 two great prothoracic horns, the minute mesothoracic ones, the very large metathoracic ones, 

 also beside the large caudal horn, the pair of small short spines on the distinct ninth abdominal 

 segment, the tubercle i of this segment being obsolete, while the suranal plate bears in its 

 middle a pah of similar spines, the plate itself ending in a long somewhat upturned conical 

 projection, extending well beyond the anal legs. Tubercles ii are wanting on the thoracic and 

 abdominal segments, at least I have been unable to detect them, while the two tubercles on 

 abdominal segment 9 are presumably ii. The two minute tubercles i on the mesothoracic 

 segment are scarcely higher than thick. Those of abdominal segments 1-7 are of uniform size, 

 being small and slender and ending in a short seta. The caudal horn is more deeply forked 

 than the thoracic ones. The suranal plate is regularly triangular, a little longer than broad. 

 The anal legs are large and rounded. The bodj r is clothed with numerous line microscopic 

 secondary setiferous tubercles. 



Stage IV: As observed by Poulton, Aglia is remarkable for passing through but four larval 

 stages. 



Length 24-25; at the end of the stage 45 mm. The stage lasts from 10 to 12 days, the 

 whole larval life in England extending from 42 to 44 days. 



The larva has now lost all its spiny armature, the shagreen dots, oblique stripes, and sub- 

 spiracular line remaining. "The whole anterior part of the larva," says Poulton, "is not unlike 

 a caricature of a vertebrate head, with the terrifying marks in the appropriate position for 

 eyes." * 



Prof. Poulton comes (1888) to the following results from his study of the ontogeny of this 

 insect. He shows the resemblance to the larva? of the Sphingidse in 15 points, also noting the 

 differences, such as the presence of the four thoracic horns and caudal horn in all but the last 

 stage. He compares the armature with that of certain Ceratocampidse, also Rhescyntis. 



He concludes that "the Sphingida? are a specialization of the group of Saturnian Bombyces, 

 and that the following order represents the nearest affinity, and is an. approach toward the 

 expression of genetic relationship: Sphinx, Acherontia, Smerinthus, Ceratomia, Lophostethus, 

 Aglia, Ceratocampa, (Attacus), Saturnia." 



[Since the above was written, Aglia tau has been made the subject of much experimental 

 research by Standfuss and others, on account of its remarkable variations, which are of great 



i The transformations of A . tau were originally described and figured by Duponchel ct Guenoe (Iconographie, ii). In 1S63, from an examination 

 of the figures, my attention was attracted to the singular changes of the larva, and I compared the young larva to the fully grown caterpillar of 

 atheroma regalis, afterwards, and pointed out that the latter genus was an "embryonic form and therefore inferior in rank to the tau moth." Amer. 

 Naturalist, June, 1S70, and Our Common Insects, p. 52. 



