424 



ELIZABETH KIXNEY. 



distinct staining reactions. The core stains a clear pink and 

 the periphery forms a black cortex around this core. This 

 reaction of the secreted substance is consistent throughout the 

 tube. 



In silk glands of D. virginica fixed in Bouin and stained in 

 iron hsematoxylin and acid fuchsin, the silk thread has a central 

 core of clear pink substance surrounded by a black peripheral 

 cortex. A wide region filled with a reticulum or meshwork of 

 fine pink granules surrounds the thread. Bordering this region 

 and extending to the inner cell-wall, a wide compact band of 

 fine granular substance arranged in concentric rings occurs. 

 Throughout this band, numerous large drops of black substance 

 appear which probably represent secretion material. By com- 

 parison of the condition just described, with that found in 

 H. cunea, and also with conditions described by Yamanouchi, 

 it appears that some difference exists between the type of gland 

 which secretes continuously and that which secretes at definite 

 periods only. The presence of two diverse staining reactions in 

 the thread seems to indicate that two substances are consistently 

 coexistent in the two different types of silk gland cell, but the 

 exact relationship between these two substances has not been 

 determined with certainty. 



(c) Critical. The question as to the identity of the substances 

 within the lumen has presented a problem which cannot be 

 definitely settled without further study both of preserved material 

 from a variety of caterpillars and also of living material. Ob- 

 servation of the material at hand seems to show that in the 

 posterior portion of the gland, the secreted substance is usually 

 homogeneous, and this is likewise true in the middle region. 

 But some modification of the secreted substance takes place in 

 the anterior conducting region which gives the silk thread the 

 appearance of being composed of two substances. 



It is evident that even in forms which secrete silk during 

 tin- entire larval life, the process is not a continuous one, for 

 often within the lumen of the silk gland, the end of one thread 

 .in be seen and, close upon the posterior part of this thread, 

 but -t [i. n.itcd from it by the substance filling the lumen, the end 

 of another discrete cylinder of secretion begins (Fig. 7). 



