Phylogeny and Evolution of the Lcindoptora. 583 



A Geometra pnpa is difficult to distinguish from a 

 Noctua, especially a deltoid pupa, and so on. 



Most species of Notodontids possess the dorsal head- 

 plate (cephalo-thoracic piece) dwindled to a small, 

 but quite definite portion. This does not at all settle its 

 position, but shows that it is below the summit level of 

 heteroceral, pupal evolution. The texture is distinctly 

 noctuid, but this has reference to the subterranean 

 position of the pupa. Centra, with a different habit, 

 might almost be Saturnian. 



The cremastral development, consisting of two or four 

 short spines, seem to be processes of the pnpa rather 

 than separate spines or bristles, as in most Nocture 

 ii some species they are quite obsolete (as in many 

 Arctians), in others (Clostera) they are at the extremity of 

 a long slender process. The chief feature in which they 

 differ from Noctuas is in the appendages tailing short of 

 the wings, but in this there is a great variety within the 

 group. The way in which the abdominal segments do 

 not materially taper till the 8th or 9th segment is dis- 

 tinctly Noctuan, but both this and the short appendage 

 covers might be claimed as Lasiocampid. 



The LiUiosiidie, Euchromiidie, and Stjntumidx are 

 hardly distinguishable from Arctiidx. The A(jaiistlds& 

 (Alyjna octomacalata is the only species I have had from 

 the e^^) is very close to Noctuida?.. Of other (exotic) 

 families and subfamilies {Pericopidse, etc.) I am very 

 ignorant. 



The Lithosiidx and some ArdiidsR [Spilosoma) are the 

 only Macro-heterocera I know whose pupte seem fairly 

 on the way to lose all movement ; in some instances, 

 possibly, they have achieved that result. This is 

 probably associated as cause and effect with their pos- 

 sessing a soft flimsy cocoon from which escape does not 

 much depend on the nature of the support and fulcrum 

 afforded by the pupal skin. 



When we come to the Bombycidstirps, there appear to 

 be very good reasons for associating together a certain 

 group of families, and these all are characterised by pos- 

 sessing a flat egg — that is, an egg so laid that the micro- 

 pyle is not on top but at one end. This egg has three 

 axes of different length, the micropylar being the longest, 

 the vertical the shortest ; it has generally a very smooth 

 surface, the sculpturing being very slight and shallow, 



