260 Annals Entomological Society of America [Vol. XV, 



The duct extends in a posterior direction and extends to the ventral 

 side of the eighth segment, where it is lost Lubben, 1907, shows that 

 these ducts pass into external, sculptured plates in the eighth segment 

 but I have been unable to find any such structure or any opening what- 

 ever at this point on any of the specimens. In H. analis the gonad 

 is oval, but appears spherical in frontal section as the long axis of the 

 gonad is at right angles to that of the body of the larva. The gonads 

 of the above three species are in pairs and lie in the fifth segment of the 

 abdomen Betten, 1901 finds the gonads for M. cinerea in segment 

 four; Lubben, 1907, records Anabolia nervosa and LimnophUiis stigma 

 as having the organs in segment five, RhyacophUa septentrionis and 

 Brachycentrus montanus in segment four Klapalek, 1888, gives the 

 gonads a general location of segment three, but neither literature nor 

 my own observations uphold this statement. From a compilation 

 of records the position appears to be in either the fourth or fifth 

 segment. 



ALIMENTARY CANAL. 



General description: The alimentary canal, in the order Trichoptera, 

 passes as almost a straight tube from buccal cavity to anus. Pictet, 

 1834, figures the tube as practically undifferentiated in its shape and 

 only varying slightly in size at the extreme ends. Klapalek, 1888. 

 states that the regions merge imperceptibly into each other. Although 

 this conc'ition may be true in a specimen gorged with food, the writer 

 has never seen a specimen where the three main divisions of stomodeum, 

 mesenteron and proctodeum were not distinctly marked one from the 

 other. The lack of convolutions as are normal in the alimentary canals 

 of insects which are herbivorous, is unusual, as Siltala, 1907, shows that 

 the order is in the main herbivorous and yet this general type of a 

 straight tube prevails. There are a few semi-carnivorous species, but 

 even here there is no variation from the above condition. The assimi- 

 lative area is increased by an unusual width in proportion to that of 

 the body, for in many places and especially the mesenteron, the width 

 is equal to one-third and frequently more than one-third the width of 

 the abdomen. Another device adds to this area or surface for the wall 

 of the mesenteron is folded into transverse ridges which increase the 

 surface by three times that actually occupied by the wall, Plate XV, 

 Fig. 10. These folds are deeper in the strictly herbivorous types, such 

 as L. indivisus, than in those which are more or less carnivorous, as 

 H. analis. We will pass now to a discussion of the alimentary canal 

 in the three species. Hydropsychodes analis Banks, Plate XV, Figs. 

 7, 8 and 9. The surface of the stomodeum of the alimentary canal 

 presents a silvery appearance, due to the peritoneal covering of the 

 muscles which surround and control the activities of this portion of the 

 canal. This portion is much darker than the remainder of the canal, 

 due to the food within. If this be cleaned, the wall appears transparent. 

 Beneath the peritoneal membrane can be seen the longitudinal muscles 

 arranged in pairs in six equi-distant places around the canal. Under- 

 neath these longitudinal muscles lie the circular muscles, which are so 



