15 



The Pterophorids in a few cases, as Pt. pentadactylus and M. 

 phatodactyla, have hairs between the scales dorsally, as in Plutella, 

 etc. In the majority the hairs are dwindled and out of sight, or 

 absent, so that the form is practically that of Obtectae. The pre- 

 sence of the lower forms, however, adds to the improbability, 

 already very great, that Pteroplwrus arises from any Pyrale or other form 

 with obtect antennae. The form in pentadactylus is a very unlikely 

 one to be reached by further evolution, and is almost certainly 

 ancestral from more generalised predecessors. 



The Tortrices have not any scales ventrally in any species I have 

 examined, and in some the dorsal scaling tends to be very narrow, 

 as in Cnephasia (zurgaweana, bellatia). The scales are always in two 

 rows ; and in the great mass of the species there is a zone of 

 sense-hairs right across the dorsum between each row of scales. 



In a small minority of species, which do not group themselves 

 together in any classification I have seen, these dorsal hairs are 

 obsolete ; the scales overlap, and there is substantially the obtect 

 type of antenna. This occurs in Tortrix (Jieparana, etc., not 

 viburnana) Teras (caudana) (not in Peroned)\ in Antithesia (soror- 

 culana, etc.) (not salicella\ Stigmonota (per/epidella, etc.), Balodes 

 (angustiorana), Ditula (ivoodiana). 



There are a few intermediate forms, such as Tortrix musculana, 

 which look at first sight as if belonging to the obtect group, but in 

 which a few hairs can be detected amongst the scales. 



Amongst the Geometers there is great uniformity throughout the 

 group. There are two rows of scales across the dorsum of each 

 segment, and hairs beneath, and there is not much difference from 

 base to apex. 



In Oporabia (dilutala) there are three rows of scales to a segment. 



In a certain number of species there is a very definite modi- 

 fication, chiefly confined to those whose males have pectinated 

 antennae. In the females of these the scales are far from being 

 definitely in lows, and they extend round the ventral aspect leaving 

 only a narrow channel for hairs, which widens out apically, and in 

 some apterous forms makes a wide brush-like pad on the seven or 

 eight last segments, in strong contrast to the nearly evanescent 

 haired surface of the proximal segments. This is the case in 

 several Hybernias. 



The species forming this group are the Boarmiadae Hyberniadae, 

 Amphidasydae and Fidoniadae, all included in Meyricks Selido- 

 semiadae. 



I have already alluded to Psodos coracina, whose antennae for 

 more than half their length apically have only one scale wide to a 

 segment, and that closely encroached on by the haired surface. 



The Noctuae have nearly the same arrangement of scales as the 

 Geometrae, viz. two rows of scales occupy the dorsal half of the 

 segment, hairs the ventral half. I have referred to the exceptional 

 condition in Anarta, 



