178 Psyche [December 



resent enzymes) and afterwards the viscid matter seems to take 

 part in the formation of the so-called peritrophic membrane. 

 The young nuclei take hsematoxylin strongly ; the secreting upper 

 portion of the cells and the cast-off disintegrating cells tend to 

 stain lightly with acid fuchsin or congo red or picric acid; or else 

 they refuse to stain at all. 



The peritrophic membrane seems to be secreted also by the 

 epithelium all along the midgut, although the nuclei of the cells 

 of the oesophageal valve appear different from those of adjoining 

 cells, as if they might be special cells secreting the whole mem- 

 brane; as concluded by Miall and Hammond in the case of Chiro- 

 nomus . *But this does not seem probable with regard to Siphanta, 

 because certain cells along the whole midgut epithelium can be 

 plainly observed secreting, some of the globules of secretion 

 adhering to and spreading over the outermost layer of the peri- 

 trophic membrane. This latter extends from the anterior end 

 of the reservoir through the whole canal to the anus, where it 

 appears to pass out with the rest of the excrement, in a granular 

 or disintegrated condition. The peritrophic membrane is seen 

 in transverse sections (fig. 4B, pm) to be composed very often 

 along some parts of its length of more than one membrane or 

 parts of several non-synchronous secretions, one within another 

 and more or less concentric. It is not very evident how the 

 peritrophic membrane itself can protect the mesenteric epithelium 

 — as it is said to do — by keeping particles of food etc. from contact 

 therewith, since the membrane is very irregular in contour and, 

 when shrunken, comparatively rough, with occasional parts of 

 cell-walls and other matters in its wall not completely digested 

 or dissolved; it is also more or less chitinous. It would appear, 

 rather, that the epithelium of the mesenteron is protected from 

 both food-particles and peritrophic membrane by the layer of 

 viscid fluid mentioned above, which is between the epithelium 

 and the membrane, and which afterwards — in part at least — 

 seems to compose the said membrane. This fluid would tend to 

 keep the latter with its contents fairly in the center of the alimen- 

 tary canal, even when rounding the numerous sharp bends of 

 the gut. But it may be that only the older internal membranes 



i The Harlequin fly, Miall and Hammond, 1900. 



