576 Dr. Thomas Algernon Chapman on the 



Can we then find any charactersof ova that will enable 

 us to determine the probable relationships of the 

 different families to each other. In studying the signifi- 

 cance of any structure, one must be led entirely by what it 

 teaches, in nowise straining the facts to meet precon- 

 ceived ideas, and not until this is done must one compare 

 the results with those otherwise attained, and then see 

 whether the differences in results are to be reconciled by 

 a fresh interpretation in the newer or older studies. 



Approaching the question of e^g form in this spirit, 

 lepidopterous eggs are seen at once to present at least 

 two very definite types of ep;g, which we may name the 

 upright and the flat egg. 



The iq^right egrj has the Ilicro'pylar axis vevixcoX to the 

 surface on which the egg is laid, and the two axes at 

 right angles to this are equal in length, and in fact not 

 definite in position, the egg being circular on any 

 horizontal section. 



The flat eijg has the MiQrojnjlar axis parallel to the 

 surface on which the egg is laid, whilst the other hori- 

 zontal axis and the vertical axis are usually of different 

 lengths. 



When we examine the distribution of these two forms 

 of eggs in different families, we find that there is rarely 

 if ever an exception to the rule, that one form only exists 

 in each family, and even in each superfamily, and that, 

 broadly, there appears a very strong presumption that, at 

 least amongst the macros, there is never an abrupt 

 transition from the one form to the other, and that whilst 

 the flat egg is probably the earliest form, and may still 

 retain a capacity for reaching the vertical form, any 

 reversion in this respect is very doubtful. 



If we confine ourselves to the true Ilacro-lejndoptera, 

 we find the upright egg occurs in the butterflies and in 

 the Noctute, whilst the flat egg is characteristic of 

 the GeometrfB and the Bombyces. 



Now I think nineteen times out of twenty, perhaps 

 ninety-nine out of a hundred, I should recognise a butterfly 

 egg from any other, but I have so far failed to find any one 

 definite character that can be predicated of all butterfly 

 eggs and will at once distinguish them from all other 

 upright eggs. Notwithstanding this, I think, we may 

 properly divide the upright egg into two divisions — the 

 butterflies and the Noctuas. 



