Meyricx.—On New Zealand Geometrina. 53 
Larva with 10 (rarely 12) legs, three pairs of abdominal legs being 
usually absent. 
A very well-defined and interesting group ; sometimes but erroneously 
regarded as a single family, though it certainly comprises several. This 
error has probably been due to Guenée, who established on superficial 
grounds families which he could not define, and thereby caused an impres- 
sion that no accurately definable families existed. I cannot pretend to any 
certainty in my views of the families hereafter defined ; they may be 
capable of further subdivision, or require partial amalgamation ; they are 
however natural and accurately limited as regards New Zealand species. 
The following considerations on the process of development of the group 
will justify the main outline of my scheme of classification. The ancestral 
form of the Geometrina must have had 12 veins in the forewings, 7, 8, and 
9 on a common stalk, the rest all separate; and 8 veins in the hindwings, 
all separate, and vein 8 free. This is the only form from which all existing 
types could have originated. Taking first the hindwings, there are two 
main types at present predominant; (A) in which there are 7 veins, 5 and 
6 separate, and 7 not anastomosing with 6; and (B) in which there are 8 
veins, 6 and 7 stalked, 8 anastomosing with 7. These form the two prin- 
cipal subdivisions of the group; it will be seen that (A) differs from the 
type by the loss of a vein (normal vein 5 which is obsolete, existing only as 
a slight fold), and (B) by the stalking of 6 and 7, and anastomosing of 8 
with 7. Probably, therefore, the hindwings of the ancestral form were 
relatively broader than in either of these, since both changes are such as 
would be likely to result from a contraction of space. This difference of 
method indicates unmistakeably that the development of the Geometrina has 
proceeded on two distinct main branches, the types of which will be found 
to correspond with the Ennomide and Larentida respectively, as I have 
defined them. Comparing the forewings of the same types, it will 
appear that in the Larentide vein 10 always anastomoses with 9, and 
11 with 10, whilst in the Ennomide they are often separate; in the 
Larentide also vein 11 often coincides wholly with 10 on lower portion, 
so that it appears to rise from the upper margin of a simple areole, or 
even from 9 above areole ; this structure, which obviously implies a 
greater remoteness from the type, is hardly ever found in the Ennomide. 
In the Larentide, therefore, the single areole resulting from this latter 
modification marks a later development, and the first three sections of 
that family are of more recent origin than the other two. So also the 
Acidalide, which have this same character, are later as a whole than the 
Larentide, from which they differ by vein 8 of the hindwings tending to free 
itself from 7, a reversionary but more recent development. Instances of 
