50 CLASSIFICATION OF INSECTS. 



and Packard and Lubbock ^ have shown that both of 

 these modifications could have arisen from the Thy- 

 sanuran type. 



The word "specialization" has different applications 

 in different minds, and is rarely used by any one natural- 

 ist with the same meaning in all cases. It is necessary, 

 in order to fix the meaning in which it is used, to select 

 some standard of reference. In our opinion this standard 

 should be a purely structural one, and not be complicated 

 with physiological considerations. While it is true that 

 the more generalized forms as a rule have simpler modes 

 of living than the more specialized, the progression and 

 degradation of types is expressed more clearly and can be 

 studied more easily in the physical structure of their or- 

 gans and parts than in any other way. Structural modi- 

 fications are also probably the direct results of changes in 

 habits and in the physical forces of the environment, and 

 are therefore reliable indications of changes in the modes 

 of life and surroundings of the animal. Thysanura is the 

 standard with which all the larval and adult forms of 

 insects should be compared. If this be accepted, it follows 

 that the primitive winged insect from which most of the 

 orders sprang was probably an animal having two equal 

 pairs of membranous wings, six nearly equal thoracic legs, 

 head distinct with biting mouth parts, and having a mix- 

 ture of structural characteristics connecting it in one direc- 

 tion with Thysanura as the ancestral form, and in still 

 others with those orders of insects which have Thysanuri- 

 form larvae. These standards of comparison enable us to 

 see that even the most generahzed groups of existing in- 

 sects are highly specialized, although, like the Odonata and 

 Blattariae, they may represent very ancient types. Speciali- 



^ Monograph of the Collembola and Thysajizira, pp. 50-52; 

 Origin and Metamorphoses of Insects, pp. 71-73. 



