106 BRITISH LEPIDOPTEHA. 



ii has not only lost its position outside i as a posterior trapezoidal, 

 but has migrated first inside i and then been moved up anteriorly 

 until in line with i, the most highly specialised arrangement of the 

 tubercles found in Psychids. The position of iii as supraspiracular, 

 iv and v as subspiracular, vi below these, and vii as marginal tubercles, 

 does not suggest any such alteration as Dyar indicates, but appears 

 quite normal.. The fact is that, in the generalised genera — Diplodoma, 

 Narycia and Solenobia — i and ii are arranged in typical trapezoidal 

 form, i nearer, ii farther from the medio-dorsal line ; in Taleporia, ii 

 has migrated directly behind i ; in Fwnea, ii has passed inside i, whilst 

 in Epichnoptevyx and Luffia it has, in addition, moved anteriorly ; in 

 the higher Psychids it actually gets in line (or almost so) with i. We 

 wish to insist on this because we have here an exact illustration of the 

 evolution of the Psychid tubercles from the most generalised to the 

 most specialised forms. Through all this modification in position the 

 tubercles maintain their original simple structure, each consisting of a 

 simple chitinous base with a single seta. 



Bruand considers that the Psychid larvae show alliance with Stygia 

 and Cochlidion, whilst Heterogynis, which Bruand and Herrich-Schafter 

 considered to be a Psychid, and to which Ave have already referred, 

 has, in spite of its evident Anthrocerid affinities, its exposed feeding- 

 habits, and absence of a case, a distinct connection Avith the Psychid 

 larva in the arrangement of its tubercles — i and ii being placed as 

 trapezoidals, iii as supraspiracular, iv and v approximating as sub- 

 spiracular, a single vii in first instar, a double vi being present, in 

 addition to these in later instars. All the tubercles in Heterogynis are 

 simple and carry a single hair. 



Psychid larvae vary much in their food requirements. Those of the 

 more generalised families are specially attached to lichens (but appear 

 to have a strong tendency to become scavengers and to affect a car- 

 nivorous diet, if occasion offer), whilst the higher Psychids, although 

 generally supposed to be attached to grasses and low plants, are some- 

 times more or less specialised to some particular food-plant. Some 

 are reputed to be almost polyphagous, and Heylaerts asserts that many 

 species, apparently specialised in a natural state to some particular 

 food-plant, will eat many other foods in captivity — lea\ r es of Ranunculus 

 bulbosus, 11. repens, leaves of salloAvs (preferably Salix alba), Calluna 

 vulgaris, Erica tetralix, Thymus vulgaris, T.sevpyllum,kc. He notes that 

 all the species of Oreopsyche (atra, plumistrella, &c.) prefer grass, Avhich 

 others also eat. The same observer states that he has entirely reared larA T ae 

 of Proutia betulina, Zell., to maturity by feeding them on dead lepidop- 

 terous imagines, Avhilst details of the cannibalistic habits of the laiwae of 

 Taleporia tubulosa [^pseudobombycella) mx&Diplodowa herminata (inargine- 

 punctella) will be found in our account of the respectiA T e species. Psychid 

 larvae moA T e freely in the sun, and AvithdraAv into their cases at night. 

 This manner of life tends to expose them to the attacks of hymenop- 

 terous parasites, Avhilst Avasps and spiders are said to kill off large 

 numbers in spite of the protection afforded by their cases. The larvae 

 of Bacotia Kcjiiiuii (tabulella) axe stated by Bruand to be, among others, 

 particularly liable to the attacks of ichneumons. Heylaerts says that 

 the genera Pezomachus, Piiujda, Hemiteles, Cremastus, and the'dipterous 

 genus Tachina, are the most persistent enemies of Psychid larvae. He 

 further states that, in 1879, he received from Milliere some female 



