BACKBONELESS ANIMALS WHICH BREATHE IN AIR 435 



lungs of backboned animals and land-snails. For, as we have 

 seen, there is marked centralization in the case of these organs, 

 impure blood flowing to them for purification, after which it is 

 distributed to the body at large. Air - tubes effect the same 

 purpose by decentralization, carrying air to and from the different 

 organs. In Peripatus the breathing organs are not very perfect 

 or specialized, but they are an early step in the evolution of a 

 mode of respiration which reaches its highest expression in 

 insects. We have no certain knowledge of the way in which 

 air-tubes first took origin, but it is extremely likely that they 

 are modifications of structures which originally served some 

 other purpose. Peripatus resembles segmented worms (annelids) 

 in several respects, and it is pretty certain that the ancestral 

 forms from which arthropods have been derived were segmented, 

 more or less worm-like creatures. It is therefore among annelids 

 that we must look for the organs which by change of function 

 have become air-tubes. Now the skin in such worms contains 

 numerous glands, by which various sorts of material are separated 

 from the blood for various purposes, such, e.g., as the formation 

 of tubes in which to dwell. And it has been suggested that 

 air-tubes have arisen from branched skin - glands which gave 

 up their original function and were specialized for carrying oxygen 

 to the body. 



CENTIPEDES AND MILLIPEDES (MYRIAPODA) 



The openings (stigmata] into the system of air-tubes by which 

 a Millipede or Centipede breathes are placed in two rows, one 

 along each side of the body. The remote ancestors from which 

 these forms are descended probably resembled Peripatus in many 

 respects, and no doubt possessed stigmata scattered over the 

 body, as well as some with a more regular arrangement (see 

 p. 434). Later on many of these were done away with, only a 

 row along each side being retained, as most conveniently situated. 

 This is another instance of the principle already illustrated by 

 the cases of gill-slits (see p. 386) and teeth (see p. 14), where 

 greater efficiency is obtained by reduction in number of structures, 

 which are in the first instance numerous and unspecialized. We 

 shall have occasion to see that among Insects the reduction of 

 these particular organs is carried a great deal further. Each 



