Phylogeny and Evolution of the Lepidoptcra. 579 



set up on end in much the same manner as in sundry 

 Geometras. 



In our common Neustria and Lanestris the eggs are set 

 up on end, so that, regarding the mere terms of the defini- 

 tion, they are vertical eggs. They have, however, three 

 distinct axes, and may be regarded as being laid not on 

 the twig at the nadir from the micropyle, but really as 

 being laid on one another, a further development of the 

 condition presented in I^ndromis and some Saturnids, 

 where the eggs are piled up on one another. 



It must be noted, however, that in some families of 

 Geometrte the eggs present a gradual development of the 

 hexagonal pitting into a regular ribbing of the noctua or 

 nymphalid type. These still retain the pitting over the 

 nadir of the micropylar area, and in the Acidalias preserve 

 three unequal axes, though in Acidalla imitaria they 

 even present a tendency to be laid on the end. These 

 are the neai'est approach to an exception to the general 

 rule that I have met with, and they suggest that the 

 upright egg is the higher form, and that nymphalid rib- 

 bing is also a late development, and that the Geometrid 

 egg is still capable of progress in these directions. 

 It would be interesting to Bud a similar progress amongst 

 Bombycid ova. 



It is therefore obvious that the two forms of eggs, as 

 we find them in the Macro-heterocera, must have been 

 derived the one from the other, or both from some other 

 form, at some much lower point in the phylogenetic 

 series. 



In trying to trace backwards the several forms of eggs,, 

 we find amongst the lower (Micro) forms a great prepon- 

 derance of groups with flat eggs. This appears to be so, 

 after making full allowance for the fact that our (or at 

 least my) ignorance of the eggs of Micro-lepidoptera is 

 considerable. The lowliest lorm with upright eggs is 

 Gossus ; this fact is curious, since both Zenzera and 

 Tortrix, to which it is allied, have flat eggs. It is pro- 

 bable, therefore, that we here have the point where the 

 two forms are still unfixed and capable of easy variation. 

 The alliance (by pupa) of Catttnia to Cossus would 

 probably point to this bemg the origin of the butterfly 

 stirps. 



But branches, by the way, are exceedingly scarce. 



Sesiidw, that I had surmised to be, perchance, a low 



TKANS. ENT. SOC. LOND. 1896. — PART IV. (dEC.) -39 



