22 HISTORY OF INSECTS. 



a few threads of silk among the leaves, or in a corner of the 

 tray. The silk is elaborated in two long slender vessels 

 lying at the sides of the stomach and intestines, and termi- 

 nating in a single tube, through which the viscid fluid, of 

 which the silken threads are composed, is forced by the pe- 

 ristaltic action of the muscles. This tube terminates in the 

 centre of the lower lip of the caterpillar. Although, how- 

 ever, the two vessels unite into a single tube, it is evident 

 that the silken threads are not united ; since we learn from 

 the recent microscopical investigations of Dr. Ure,* that 

 each of the silk threads was found to be composed of two 

 distinct cylinders, which in good silk appear to be perfectly 

 parallel and quite cylindrical; the unevenness or the want 

 of parallelism producing inferiority in the raw material. 

 Each of these cylinders varies in diameter from one two- 

 thousand two hundredth part of an inch (the measure in 

 silk of the best quality), to one eighteen hundredth of an 

 inch. When imported, however, several distinct threads 

 are found to be reeled together, which is done by the grower 

 of the silk- worms, the threads being passed through several 

 eyelets, and then wound off. The imported raw silks are 

 of various qualities, depending upon the mode and time of 

 feeding, as well as upon the food of the silk-worms. These 

 compound threads are one five hundredth part of an inch in 

 diameter, each being divisible into eight threads ; that is, 

 into four pairs of cylinders, in the best silk. The specific 

 gravity of silk, according to this author, is greater than had 

 been generally supposed, being 1256 to 1200, and conse- 

 quently greater than the strongest muriatic acid or water. 

 When placed in the paper cell the silk -worm spins a few 

 loose connecting threads, attached to every side of the cell : 

 it then forms a regular oval ball of silk, which is suspended 



* ' Trans. Ent. Soc.' vol. i. Journ. of Proceedings, p. 50. 



