194! Ur. T. Williams on the Mechanism of Aquatic 



investment, will appear, from the close and tubulai* manner in 

 which it is embraced by this membranous investment, to throw 

 off, as supposed by Mr. NcAvport, its external coat at the point of 

 entrance. Acetic acid however proves this appearance to be false. 

 The tube still preserves its three constituent elements after enter- 

 ing the substance of the organ, whatever it be, and until it 

 assumes the capillary or membranous character. 



It is important to observe, because it reconciles the accurate 

 observations of INIr. Bowerbank with those of the author, that on 

 the wings (fig. 10, f), especially on the scaly intervals between 

 the nervures, the spiral tracheae, as correctly stated by Mr. Bower- 

 bank, do not for the most part degenerate into the membranous 

 tracheae. In these situations the spiral continues to the extreme 

 termination of the tube (fig. 11, 6). There is something ana- 

 tomically characteristic in the walls of the membranous tracheae. 

 They denote a difference between those ' parietes ' through which 

 a gas has to pass, and those (of the vessels) which fluids transude. 

 The smallest trachea differs from the largest only in the absence 

 of the spiral, just as the largest artery differs from the capillai-y 

 only in the presence of a thick elastic coat. The tracheae ter- 

 minate differently, and form different plexuses, in different organs, 

 according to the varying mechanical arrangements of the ulti- 

 mate parts of the latter. 



The conclusion must be emphatically reiterated, that however, 

 wherever, and in whatever structure the tracheae may periphe- 

 rically terminate, they are «ir-tubes throughout all changes to 

 their final extremes. 



Distribution and Subdivision of the Trachea. 



The primary, secondary and tertiaiy tubes divide and sub- 

 divide arborescently, the branches never reuniting (fig. 14). In 

 the spiral tracheae no plexiform union of the branches ever, 

 anywhere, occurs ; so far the observations of Mr. Bowerbank and 

 Mr. Newport are exact to nature. It is because these distin- 

 guished observers could not succeed in tracing the air-tubes 

 beyond this limit, and because they drew a general inference 

 intended to be applied to all structures, from the distribution 

 and termination observed by the tracheae in the wings, that they 

 were both seduced into the error of supposing first these points 

 to be the distributive ultimata of the tubes, and secondly, that 

 the tracheae nowhere inosculate. As already stated, this is true 

 only of the scaly intervals which separate the nervures of the 

 wings — of no other structures. In nearly eve}-?/ other structure in 

 the body of the Insect the air-tubes divide and subdivide in the 

 same profuse retiform manner as the blood-capillaries of the 



*8 



