EVOLUTION 327 



In seeking to trace the course of the development of 

 insects as a class we have to compare the characters of the 

 orders and families as regards structure and life-history so 

 as to determine which groups are the more primitive and 

 which the more specialised. Among all classes of animals 

 there are certain groups which seem clearly to have changed 

 less in the course of time than others. It would obviously 

 be rash and probably quite wrong to infer from this that 

 the more specialised groups are direct descendants of the 

 less speciahsed. The fact that both grades are living 

 together in the world to-day suggests that they are collateral 

 descendants of forms now extinct. But we are justified in 

 concluding that the more primitive of surviving insects 

 resemble more closely an ancestral stock, more or less re- 

 mote, than do members of the more speciaHsed orders. 

 The survey of insect classification in the previous chapter 

 indicates the wingless Apterygota as the most primitive 

 of the sub-classes. As insects are the only members of the 

 Arthropoda endowed with the power of flight by means of 

 wings, it is certain that the common ancestral stock of 

 insects and other classes of arthropods were wingless 

 creatures, and that the early insects before they acquired 

 wings had assumed the characteristic insectan distinction 

 between the thorax with its three pairs of legs and the 

 abdomen with its limbs mostly reduced or wanting ; the 

 dominance of the thorax as the locomotor centre of the 

 insect seems a necessary preliminary to the development of 

 wings in that region of the body. We feel confident, there- 

 fore, in regarding such a bristle-tail as Machilis as resembling 

 the primitive stock of the Insecta more closely than any 

 other living type, but we are not warranted in referring to 

 the unknown members of that stock any detailed characters 

 of Machilis except those that are clearly indicated as 

 primitive. The body segmentation with its three thoracic 

 and eleven abdominal segments, the crustacean type of 

 mandible, the superlinguae, the abdominal appendages, 

 reduced except the long tail-cerci, all these are thysanuran 

 features which may confidently be claimed as primitive 



