124 



THE ENTOMOLOGIST S RECORD. 



the head of another phylum. In losing its scales, Hepialm retains 

 some dorsal hairs which it probably had never lost. Saturnia, Danais, 

 &c., have lost the dorsal hairs, and so the scaloless surface is naked, 

 except where pits are partially regained by lateral encroachment. 

 Dr. Jordan regards the scaleless dorsum of the last joints of many 

 butterfly antennae as remains of the ancestral scaleless antenna. In 

 view of the facts exhibited by the lower Pltalacnae, this is by no means 

 certainly the case. It appears to be rather a further evolutionary stage, 

 resulting from a day-flying habit. 



In several day-flying I'halcwnae, the tendency to distal disappearance 

 of scales is very strong. In Ai/aristtdac, c.;/., Ahjpia octtniiacidata, in 

 such Noctuids as Anarta, or such Geometrids as Psodus, the dorsal 

 scaling is reduced to a line of two, and finally of one, scale, and for 

 several terminal joints there are no scales. Where, as in several 

 instances, I have counted more than four or five such joints, I have 

 suspected that the solitary remaining scale may have been removed by 

 accident in some segments. Even assuming that it existed to the tip 

 (and my whole observation, so far as regards absolute scalelessness, is 

 due to sucli scales having been accidentally removed), the paucity and 

 weakness of scaling towards the tip of the antennae in these forms, 

 remains as a very strong and definite fact. Why other day-flyers such 

 as Syntuuiis, Stasia, and Anthrocera, should not equally tend to have 

 scaleless tips dorsally, is not very clear. Hesjienidae are in the same 

 case, but some species evidence the commencing change, which has 

 not perhaps had time to take effect in the Sesiids, and others. Si/ntuniis 

 and Anthrocera, are notable for a " booming " flight, i.e., as I take it, 

 a straightforward and slow flight, uncomplicated by any rapid changes 

 in pace or direction ; whether this accords with the simpler scaling 

 may be worth considering, robjplova ridois with the last three- 

 fourths of its antenna covered with sense-hairs, and yet certainly 

 descended from moths with scaling dorsally to the tip, is perhaps an 

 even stronger case. Nevertheless, Dr. Jordan's conclusion that Lycaenids 

 originated at a very low point, possibly near Jiii/atae, is probably more 

 correct than the reasoning by which he arrives at it. There is much 

 reason to believe that Lycacnidae did not proceed from Hesperids, but 

 from a common ancestor a little further back, in which the terminal 

 antennal joints probably possessed dorsally both hairs and scales, as 

 in Kriofrania {jmrjinrella) and in lower Xe(dejiid<ijitera. On Hesperids 

 the hairs vanished, the scales persisted ; in Lycaenids the scales 

 disappeared and the hairs persisted. My point is that the ancestral 

 Lycaenid had scales on the dorsal aspect of the terminal joints and, in 

 fact, all over the ventral surface also, but I see nothing to prevent 

 the hypothesis being held that the dorsal hairs have maintained 

 their position right through, from the scaleless antenna of the non- 

 lepidopterous ancestor, and have not necessarily migrated from below 

 (the Hesperid position), supplanting scales. Tmt if so, scales were 

 certainly once associated with them. 



In C/iiiii(d)arrJte {fat/cUa) the scaling is of ordinary obtect type. 

 Under this name. Dr. Jordan correctly describes the scaling of the 

 antennae of the long-horned Adelidae. It is unnecessary, however, 

 to deal more fully with Phahenal antennae, the main point for the 

 present is that in the early lepidoptcrous antenna both scales and hairs 

 occurred on all its surfaces, and that their segregation was an after 



