450 



debris (fig. 88). Indeed, it is difficult to state how much 



, . of the disintegrated tissue is removed in this manner, and 



how much in the more direct way of solution in the blood 

 stream. 



Of the occurrence of phagocytosis of the gland cells, how- 

 ever, there can be no doubt whatever. Indeed, the salivary 

 glands, in their degeneration, offer as clear an example of 

 ~ phagocytosis as it is possible to wish for. But their death, 

 degeneration, and even partial disintegration previous to 

 I phagocytosis are equally clear. 



The bifurcated portion of the salivary ducts disintegrates 

 at about the same time. The leucocytes, having removed 

 the (^ehris from the glands, now move forwards and absorb 

 the ducts also, and in the pupa six hours old no trace, of the 

 larval glands is any longer to be recognized. 



The metamorphosis of the insect intestine has been the 

 subject of a number of distinct investigations, to which I 

 can refer but briefly here. Weismann, Kowalevsky, Lowne, 

 and more recently Perez (1910), have carefully examined the 

 process in Calliphora, and it seems to differ but little from 

 that of Nasonia, soi far as essential characters are concerned. 



Deegener has investigated the metamorphosis of the 

 intestine in the Coleoptera Hydrophilus (1910), Cybister 

 (1904), and in Malacosoma (1908); while Rengel has made 

 observations on Tenebrio molitor, and several water-beetles. 



The metamorphosis of the intestine of the silkworm has 

 been investigated by Verson (1898, 1905); Poyarkoff (1910) 

 examined that of a Chrysomelid beetle Galeruca; and Russ 

 (1908) studied it in the Trichoptera. 



The observations of these workers differ considerably, 

 and while differences in the material dealt with may account 

 in part for the discrepancies, misrepresentations must not be 

 forgotten. Thus, while the formation of a replacing epithelium 

 in the midgut, confined to the pupal period, and similar to 

 that occurring in Nasonia, is fairly frequent, it appears to 

 be absent in some forms. For example, Deegener could not 

 observe it in Malacosoma; in the Trichoptera Russ failed to 

 observe it, and regarded a constricted part of the imaginal 

 midgut as functioning in its place. Verson was not able to m 

 see it in the silkworm; since its function is to absorb the 

 products of degeneration of the larval midgut, its absence in 

 the silkworm may be correlated with the voiding, as observed 

 by Verson, of these degeneration products through the anus, 

 shortly before pupation. 



If the observations of Verson are correct, the process offers 

 a curious type of inefficiency — the waste of certain useful 

 storage materials — which does not occur in Nasonia. In 



