386 BRITISH LEPIDOPTERA. 



cases, it tends rather towards the Amorphid than the Saturniid 

 shape, as typified by Saturnia pavonia. Some pupa-cases of a 

 large South African Saturniid moth that I have received are very 

 similar in shape to the pupa of S. pyri, with the exception that the 

 abdominal segments 5, 6, 7 arid 8 are ridged at the lower edge, 

 and the cremaster is reduced to a blunt point, with a very stout 

 pyramid-shaped projection on it. In fact the cremasters of this 

 unnamed species, of ,S. pyri and of Dimorpha versicolora form a very 

 good series of connecting links between the anal spike of the 

 Sphinx pupa and the bristly cremaster of S. pavonia, the order 

 being — 5. pavonia, S. pyri, D. versicolora, the South African 

 Saturniid, Smerinthus, Manduca." 



Bacot, as shown above, looks upon Mimas as exhibiting more 

 generalised characters than the other Palaearctic Amorphids, of 

 which he considers the Sichiid group the base • from an ancestral form 

 near this latter tribe he concludes that the more specialised Amorphids 

 arose, splitting up into at least three distinct groups — the Smerinthids, 

 Clarkiids and Amorphids (sens, strict.), so that the division of the 

 subfamily Amorphinae, so far as the Palaearctic and some Nearctic 

 genera are concerned, would appear to work out as follows : 



1. Mimantidi — Mimas [tiliae). 



2. SlCHUDi — Sichia* (guercus), Burrowsia (roseipennis), Kayeia (mackii). 



3. Smerinthidi — Daddia [kiiidermanni), Bellia (caecus), Smerinthus 



[ocellata), Nicholsonia [saliceti), Calasymbolus {astylus). 



4. Clarkiidi — Clarkia [dissimilis). 



5. Amorphidi — Triptogon (modesta), Amorpha ' populi). 



The following comparisons of the eggs, larvae, pupae, and 

 certain imaginal structures of our three British species, Mimas tiliae, 

 Smerinthus ocellata and Amorpha populi, should prove interesting 

 and instructive : 



Ovum. — The eggs of the three species are very similar ; they are all egg-shaped 

 and of the flat type ; they are laid on their flat sides, those of Amorpha and 

 Smerinthus are fixed by a certain amount of gluey material that is not obvious on the 

 other surfaces of the egg ; those of Mimas are more densely covered with an india- 

 rubber-like gum ; all are pale green in colour, those of Mimas are darker, either 

 actually or from the effect of the gummy coating-. All are about 2-00mm. in length, 

 those of Amorpha and Smerinthus fractionally above, of Mimas below, that length. 

 In width, that of A. populi is the broadest, and makes the nearest approach to a 

 spherical egg, its other diameters being about 17mm. and [-6mm. respectively, that 

 of S. ocellata has these 1*5 5mm. and 1-3111111., and those of M. tiliae 1-55111111. and 

 I -35mm. These measurements are. however, only roughly correct or are rather the 

 maximum diameter, and are measured nearer the micropylar end, and all taper to the 

 nadir, whether as seen from the top or from the side. Taking- a diameter at the 

 same distance from the nadir that the longest is from the micropyle, the c.^ of A. 

 populi, seen from above, measures 1-5111111. against maximum of 1-7111111., and from 

 side 1 -45mm. against 1 -60111111.; those of S. ocellata and M. tiliae, being narrower 

 eggs, differ inappreciably as seen from above, but seen from the side the measure- 

 ments are: A. populi — greatest height 1 -6111111.. height nearer nadir 1-45111111.; S. 

 ocellata, 1-3111111. and 1-15111111., and M. tiliae 1-35111111. and riomni. respectively; the 

 egg of M. tiliae, therefore, tapers most in a side view. The surface sculpture of all is of 

 hexagonal cells, tolerably regular about the equator, but towards the micropylar end 

 becoming somewhat lengthened parallel with the length of the egg. The cells are 

 about -02111111. in diameter, perhaps a shade larger in A. populi. In A. populi the 

 lines dividing the cells are line, and the centre of each cell has a small raised boss 



* Grote robbed us of Lao/hoc in 1874 by fixing populi as the type of the 

 genus, and thus annulling it, as a synonym of Amorpha ; later, in [895, he gives us 

 (/in reus as the type of Polyp/ychiis, but Stephens had, in 1850, made populi the 

 type of /'olyp/ychus, which also made this genus a synonym of Amorpha, 



