NO. 6 CATERPILLAR AND BUTTERFLY — SNODGRASS 33 



and ovaries are present at hatching, and are in the process of dividing 

 into compartments. No considerable change occurs during the first 

 four larval instars, but further development proceeds during the rest 

 of the larval life, including formation of the outlet ducts. 



From these examples of imaginal development during the larval 

 stage it is evident that an inhibitory influence does not equally affect 

 all adult structures. Bodenstein (1957) has given some attention to 

 this phase of the subject. He notes that in the larva of Drosophila 

 the imaginal discs (i.e., rudiments of imaginal organs) grow at a 

 rather constant rate, as shown by Eassa in Pieris. The growth of 

 these rudiments during the intermoult periods indicates that the con- 

 centration of the thoracic-gland hormone is sufficient for the growth 

 of these parts, but is not enough to induce moulting. "From all this," 

 Bodenstein says, "one must conclude that the ability of different 

 tissues to respond with growth to a given titer of prothoracic-gland 

 hormone varies. Some tissues are able to grow in a low, others only 

 in a higher titer." 



The thoracic-gland hormone brings about moulting, but the develop- 

 ment of the next larval instar and whatever differentiation it may make 

 is controlled by the corpus-allatum hormone. It would seem, there- 

 fore, that the whole developmental process that takes place in the 

 larva depends on the relative concentration of the two hormones in 

 the blood. The way it works out seems to be that development is 

 permitted in adult tissues and organs whose growth does not interfere 

 with the functions of the larva, while the purely larval structures are 

 maintained and larval growth is allowed to proceed. 



The body wall, the alimentary canal, the muscular system, the 

 nervous and circulatory systems, the tracheae are all preserved intact 

 to the end of the active state of the caterpillar. With species that spin 

 a cocoon and remain quiescent within it a varying length of time before 

 pupation, metamorphic changes may begin during or shortly after 

 spinning. In the silkworm, Bombyx mori, Verson (1905) says that 

 degeneration of the mesenteron epithelium begins after the emptying 

 of the alimentary canal at the time of spinning. As the larval cells are 

 thrown off they are replaced by regenerative cells, so that the mesen- 

 teron wall remains continuous. It is observed likewise by Deegener 

 (1908) that metamorphic changes in the alimentary canal of Mala- 

 cosoma castrensis begin in the spinning larva. In Vanessa, according 

 to Henson (1929), degenerative changes in the mesenteron take place 

 in the fifth larval stage. In Heterogenea limacodes it is said by Samson 

 (1908) that the larval epithelium of the mesenteron degenerates and 

 is replaced by the imaginal epithelium even during the spinning by the 



