LUFFIIDAE. 231 



tions of this sex, given under each species. The Luffiid females are especi- 

 ally characterised hy the highly -developed scales that surround the 

 abdominal segments. The comparison of the scales presented by the 

 females of this and the allied families has been worked out by Chap- 

 man as follows : 



1. In L. lapidella, ? , the scales cover the whole abdomen, except the anterior 

 border of the 1st segment, down to the commencement of the anal wool ; broader 

 forward and dorsally, they are yet of more uniform size and form throughout 

 than those of L. ferchaultella, triangular, broad at tip, narrowing regularly to base, 

 extreme breadth one-fourth to one-fifth of length, and cut off rather squarely at 

 end without any definite toothing. 



2. In L. ferchaultella, ? , the scales are very variable in form and size, not only 

 do those on examples from different localities vary, but those from the same locality 

 also show great differences in different individuals, some presenting scales very 

 like those of L. lapidella, others extremely broad ones. The broadest are usually 

 dorsal and anterior, but the very broadest are sometimes lateral. On one specimen 

 the scales are not very broad, and are rounded as in S. lichenella. In most they 

 are much like those of L. lapidella, but broader, and with various irregular tooth- 

 ings at extremity. In a few the scales (or some of them) are so broad as to be fan- 

 like, or like the leaves of Salisburia adiantifolia, broader at the extremity than they 

 are long, but preserving the straight sides, so as still to be triangular in shape, the 

 extremity somewhat irregular, with toothings that look rather as if the ends were 

 broken off than trimly serrate as in most toothed scales. 



3. In B. sepium, ? , the scales are a shade larger, and narrower in proportion 

 to width, than those of L. lapidella, usually rounded at end, or with one indenta- 

 tion, and rather more variable. 



4. In Solenobia ? and Taleporia ? the scales are more variable — extending 

 from scales to hairs — narrower than in L. lapidella in most species (some broader 

 on T. tubulosa), all that are obviously scales narrow somewhat to tip, and are there 

 rounded. 



It would appear that the tendency for the chitinous plates of the 

 $ Luffiid abdomen (especially dorsally) to be divided into two by a 

 median division, is characteristic of this group and does not appear else- 

 where. It is so far developed that in J B. sepium there appear to 

 be two dorsal and one ventral plate, in ? L. ferchaultella two dorsal and 

 two ventral. 



There is one species (generally referred to this group) that must be 

 considered obsolete, viz., the Psyche lichenwn of Schrank (Fauna Boica, 

 ii., p. 92), and referred to by Zeller (his, 1838, p. 718 and Linn. Ent., 

 vii., p. 358). This species is founded on the following description : 



Wohnort : an Eichen und Fohrenstammen, wo die Raupe ihren Sack mit den 

 Trummern kleiner Schuppenfiechten und Warzenfiechten uberkleidet, denen sie 

 auch nach der Lange gelegte diinne Splitter oder Trummer von Fohrennadeln 

 einmengt. Den Schmetterling sah ich nicht (Schrank, Fauna Boica, ii., p. 92). 



It is difficult to know what species can be said to " cover its case 

 longitudinally with thin chips or fragments of pine needles," and 

 Schrank's identification of Reaumur and Fuessly's species as cospecific 

 does not make the matter any clearer. Reaumur's pi. xv., figs, viii and 

 ix, are distinctly Luffiid, one might even suspect them to be bad figures 

 of the cases of sepium ox ferchaultella, but the description of the larva 

 that is said to inhabit them disagrees entirely, for all the known 

 Luffiid larvae are very dark, whereas this is said to be "yellow or 

 greenish-white." Fuessly's Archie, ii., pi. xii., figs. 8-9, appear refer- 

 able to 1'. bctulina, and might even represent an immature case of 

 Sterrhopterix hirsutella (calvella), as Werneburg suggests. It appears 

 to us, therefore, that Psyche licJienwn, Schrank, cannot at present be 

 referred to any known species. 



