THE ANATOMY OF THE ADULT MOSQUITO 103 



side projects into the depression formed between two con- 

 secutive cells of the other, so that in a tube that has 

 been a little flattened out by pressure, the lumen appears 

 zigzag, and the individual cells roughly triangular. 



The last part of the digestive canal is much narrower 

 than the stomach, its lining epithelium smaller celled and 

 of a less glandular character, and its wall thick and mus- 

 cular, while around it may be distinguished some strands 

 of voluntary muscular fibre springing from the wall of the 

 penultimate segment. 



By a little judicious teazing of the last two abdominal 

 segments, there may be made out, lying on the ventral 

 aspect of the hind part of the intestine, four pear-shaped 

 follicular glands, richly supplied with a close net-work of air 

 tubes, and connected with the intestine by short ducts. 

 These are lined with large gland cells, and probably secrete 

 some fluid accessory to digestion. In the natural position 

 they are arranged not symmetrically in pairs, but are 

 " echelonned " in obliquely placed couples. 



The respiratory system is, for the most part, quite of 

 the usual type, the main trunks from the stigmata giving 

 off communicating branches to those in front and behind 

 them, and ending in a tuft of branches for the supply 

 of the muscles and other organs contained within the 

 segment. There are no distinct stigmata in the head or 

 in the last abdominal segment, though I have observed 

 certain tracheae undoubtedly communicating with the ex- 

 terior near the base of the proboscis ; and the two 

 anterior thoracic stigmata are by far the largest and 

 most important in the body, the abdominal openings being 

 individually small and difficult to make out, as they are 

 completely hidden by the scales fringing the edges of the 

 sterna and terga. They are placed rather nearer the front 

 than the back of the segments, and are best demonstrated 

 by first carefully removing all scales by brushing, and then 

 crushing the insect as it lies on its side between cover and 

 slide, so as to pinch the abdomen from side to side and 

 bring the lateral surface into view beneath the microscope. 



There is, however, in this connection, one structure that 



