276 BRITISH LEPIDOPTERA. 



bosses are slender, short, very minute hairs, originating from a swollen base, athe- 

 roma regalis : The polygonal impressed cells readily recognised under micro- 

 scope ; much more distinct and crowded than in two preceding. 



3. Hemileucid^e. — Hyperchirict io : Finely granulated, not smooth and 

 shining ; shows (under ^ in. objective) the. surface divided into closely-set, very small, 

 slightly raised but flattened areas, separated by narrow valleys ; the areas very 

 irregular, but somewhat polygonal in outline. 



The typical Attacid larva tends to have the tubercles arranged 

 in circular form around each segment, this arrangement being brought 

 about by the atrophy of ii, leaving i, hi, iv + v, and vii as raised 

 warts on either side of each segment, bearing, in many cases, power- 

 ful spines or remarkable hairs. But it is in this stage that the 

 superfamily appears to attain its highest form of specialisation, and 

 the modifications are so extreme, that one finds within its limits 

 perfectly smooth larva?, as well as some of the most hugely spined 

 of all lepidopterous larvae. The suranal plate is, in some larvae, 

 exceedingly developed, e.g., in Aglia tau it forms a long, permanent, 

 sharp spine. In the Citheroniids — Anisota, Dryocampa, Eacles and 

 Citheronia — it is very large, the surface and edges being rough and 

 tuberculated. It appears to attain its maximum in Sphingicamp'a, 

 being triangular and ending in a bifid point. The 9th abdominal 

 segment is occasionally well developed in the Attacids and Cither- 

 oniids, sometimes bearing a true caudal horn, which replaces that 

 more frequently developed on the 8th abdominal segment (Packard). 



One finds that when the young larvae of some of the spined 

 forms leave the egg, the spines are flaccid and soft. Packard observes 

 ( Proc. Am. Acad. Arts Science, 1893, p. 59) that, when the young 

 larva of Platysamia cecropia hatches, the yellowish-green tubercles are 

 flattened close to the body and the hairs or setae in each verticil or 

 pencil are united in one pencil-like mass and bent to one side of the body. 

 In ten minutes the tubercles become erect, higher and longer 

 (probably swelled out by the pressure of the blood), and, by this time, 

 the hairs have assumed their radiate arrangement, and in one or two 

 minutes more (viz., 11 — 12 minutes after leaving the egg), the tubercles 

 have all become of full length and erect and the black setae spread out 

 in the normal verticillate way. He further notes (loc. cit., p. 87) that 

 when the larva of Act/as luna is leaving the egg and before entirely 

 breaking out of the eggshell the tubercles on the anterior segments 

 become erect, and the hairs radiate from them, but, behind, the 3rd 

 thoracic and abdominal segments are seen to be soft and flattened or 

 appressed to the body, and adhering in flaccid bundles. In P. cecropia 

 on the other hand, all the tubercles and bristles are flabby for perhaps 

 half-an-hour after the creature frees itself from the egg ; ten minuter 

 after hatching all the tubercles of one were seen to have become filled 

 out and erect with stiff radiating bristles. It is evident that before, 

 and at the point of, hatching the setoe or bristles are filled with blood 

 which distends them. While thus distended the fluid may ooze out 

 of the ends and thus they may be called glandular hairs. In those 

 which are full and bulbous at the end, the fluid may be retained 

 through stage 1, and, in rare cases, through the second or even the 

 third stage. See also Senile, Psyche, ix., pp. 299 — 300. on the develop- 

 ment of the spin}- tubercles in newly-hatched larwe of Eacles imperialis. 



Many Attacid larvae, e.g., Citheronia rcgaiis, Automeris joints, See, 

 are no doubt protected by their spines, others, e.g., Telea poiyphemus, Actios 



