in the pupe of Heterocerous Lepidoptera. 118 
related to the very unusual immense ventral flange on 
that segment and the great dwarfing of segments 8, 9, 
and 10. These and all the other characters of Hepialus 
would lead one to place it very distinctly from Cossus, 
yet there appear to exist in Australia many forms uniting 
Cossus, Zeuzera, and Hepialus into one family, both in 
their imaginal and pupal characters, and it must be in 
Australia that the derivation of a Hepialid form from an 
Adelid one, and of Tortrix from the Cossus side of the 
group, must be worked out. No doubt intermediate forms 
do or did exist between any two families that may be 
named, and therefore that we know many of them in 
any particular case, as that of Cossus and Hepialus, 
should not prevent us regarding these as tolerably wide 
apart. 
The metathoracic structure of Hepialus came as a very 
unexpected confirmation of the idea that of the Tortri- 
coid group it was the nearest to the lower Adelids, and, 
despite its specialisation, was near the line by which 
Tortrix was derived from some Adelid form. 
Since the removal of Cossus, &c., from the Bombyces 
to the Micros will have to meet a great deal of what has 
become almost instinctive belief, I may mention one or 
two further points supporting the Micro character of 
these families. To take, first, the pupa: there is in the 
pupal outline of the Micros (Incomplete), viewed laterally, 
a strong tendency to form a waist by the sinking in dor- 
sally of the 5th segment (1st abdominal) ; this is almost 
unknown in the Macros (Obtecte). This is well seen in 
Cossus (as in all Tortrices). Bembeciformis presents this 
character markedly. 
The wing-patterns of all (? Arctia, &c.) true Macros con- 
sist of transverse lines in certain well-known positions, 
and of certain stigmata. These do not exist very dis- 
tinctly in many of the Pyraloids, except, perhaps, true 
Phycids, and are wholly wanting in all Incomplete. 
The Adelide may be taken as showing the fundamental 
pattern in these, a uniform colour strongly tending to be 
metallic, and with markings in irregular patches and 
spots (L. rubiella) ; the more definite pattern of L. capi- 
tella also obtains largely amongst Micros. Now, none of 
these families—Zygena, Cossus, Sesia, &e.—present the 
Macro pattern; Zygena, the highest of them, has even 
retained the metallic tints of Adela, and is marked by 
TRANS. ENT. SOC. LOND, 18935.—PART I, (MARCH.) TI 
