STATES OF INSECTS. 147 



stripes, in which these horns are of equal thickness at 

 base and apex, but with the same terminal knob. Da- 

 nais Archippus has a pair of tentacula at the head, and 

 another pair, but shorter, at the tail ; and D. Gylippus 

 has, besides these, two in the middle of the body a . 



We are equally ignorant of the use of the upright horn 

 found upon the back of the fourth segment in the larva of 

 some moths (Noctua Psi, and tridens F.) which is of a con- 

 struction quite different from that of those last described. 

 It is cylindrical, slightly thinner at the apex, which is 

 obtuse, fleshy, incapable of motion, of a black colour, and 

 about two lines long. On the same segment, also, in the 

 case-worms ( Trichoptera K.) are three .fleshy conical emit 

 nences, which the animal can inflate or depress, so that 

 they sometimes totally disappear, and then in an instant 

 swell out again. When retracted, they form a tunnel- 

 shaped cavity, varying in depth b . Reaumur conjectured 

 that these eminences were connected with respiration, 

 and one circumstance seems in favour of this conjecture, 

 that this segment has not the respiratory threads observ- 

 able in the subsequent ones. Latreille mentions certain 

 fleshy naked eminences placed upon the ninth and tenth 

 segments of some hairy caterpillars, which, like those just 

 mentioned, the animal can elevate more or less. They 

 are often little cones ; but when it would shorten them, 

 the summit is drawn in, and a tunnel appears where be- 

 fore there was a pyramid c . 



In a former Letter I gave you a short account of the 



a Smith's Abbott's hisects of Georgia, t. xiii. 

 b De Geer ii. 507. t. x\.f. 16. m n. t. xiv./. 7- 

 c N. Diet, a" Hist. Nat. vi. 256. 



L 2 



