1916] Reflex " Bleeding'' of Epilachna 209 



the ampulla is generally indiscernible and no tubes were seen 

 running into the cells, although it is common to see the free 

 ends of the tubes among the cells. That no tubes were seen 

 entering the cells is easily explained by supposmg that they 

 were pulled out of the cells by the hypodermis withdrawmg 

 from the chitin during the fixation. The injury produced by 

 the tubes being pulled out of the cehs is certainly the cause of 

 the ampullae being rarely seen. While examining live gland 

 cells in the honey bee, the writer (1914) found that the slightest 

 injury to the cells caused the ampullae to disappear. 



At certain places the gland cells (Fig. 21, glc^) lie bead-hke 

 in the hypodermis, causing the hypodermal wall to bulge out 

 around each of them, while at other places they He m groups 

 (Figs. 18, 19 and 21 glci) among the cells belonging to the other 

 type of hypodermal glands. 

 (1) Origin of hypodermal glands without reservoirs. 



Since no material of pupae was prepared, Httle can be said 

 about the origin of these glands, because the conducting tubes 

 are fully developed and the gland cells are being rapidly formed 

 when the insect emerges. When the insect has been emerged an 

 hour, large hypodermal nuclei (Fig. 19, hnuc) may be seen 

 beneath the articular membrane, and many small cells, pre- 

 sumably gland cells {glc), he in and below the hypodermis 

 among larger gland cells iglc). 



There is no doubt that these gland cells are modified hypo- 

 dermal cells, but since the conducting tubes pierce the entire 

 thickness of the chitin, they must have been formed simul- 

 taneously with the chitin. The conducting tubes belonging to 

 hypodermal glands of the same type in the honey bee are 

 formed as follows: Just before the hypodermal cells begin 

 secreting the chitin, some of them begin sending out processes 

 into the lumen of the abdomen. At the same time other hypo- 

 dermal cells migrate into the lumen and the processes sent into 

 the lumen unite with the cells that have migrated inward. 

 Each process forms a chitinous tube which runs from the ampulla_ 

 to the surface of the chitin. By the time the chitin is fully 

 developed, the gland cells are also fully developed, the proto- 

 plasm in the processes has been withdrawn from around the 

 tubes, and the tubes are firmly embedded in the chitin. 



