HOW INSECTS BREATHE. 137 



mesothorax and metathorax ; they carry the three pairs of 

 legs, and, in the case of the two latter, also the wings, and 

 are in the perfect insect more or less intimately connected 

 together, forming a sort of strong box called the thorax. 

 The remaining nine rings form the abdomen, but are very 

 seldom all of them developed. 



The normal situation of the spiracles is on the membrane 

 connecting together each pair of segments, except the first 

 and last. There are, therefore, never more, and very often 

 less, than ten pairs, depending on the number of developed 

 abdominal segments. 



In a great many cases, however, the spiracles are situated 

 not on the connecting membrane but on the segments them- 

 selves. In the silk worm, for instance, or the large caterpillar 

 of the Goat moth, in which the spiracles can be seen as a 

 row of nine dots on each side of the body, the spiracles 

 which, according to the above rule, should be between the 

 pro and mesothorax, are on the prothorax, while those which 

 we should expect to find between the metathorax and first ab- 

 dominal segment are carried back to the first abdominal se°;- 

 ment and so on with the posterior spiracles. We cannot, 

 however, at present account for these slight variations of 

 position, and this is not the place to describe them in detail. 



In the human throat there is a contrivance by which 

 foreign substances are prevented from passing down the 

 windpipe into the lungs ; it is simply that the windpipe is 

 smaller during expiration than in inspiration, the current of 

 air rushes out with greater velocity than it entered, and of 

 course carries out with it anything that may have got in 

 accidentally :* in insects there is no such contrivance. The 

 reason of this difference is obvious. Suppose in man, for 

 instance, we had a complicated apparatus to prevent any 

 of the food from entering the windpipe, but no means for 

 * Sir C. Bell, Phil. Trans. 1832. 



