SYSTEM OF INSECTS. 375 



ideas on this subject in a former letter, I need not here 

 repeat what I then said a . 



iii. As we have subkingdoms, so we may also have sub- 

 classes, or such large divisions of a class r-not founded up- 

 on internal organization or any of the primary vital func- 

 tions, but upon different modes of taking their food, or 

 such other secondary characters as include more than 

 one Order. To this description Clairville's Mandibulata 

 and Haustellata appear to me to belong, which I think 

 are by no means entitled to the rank of Classes; for 

 whoever compares these two tribes together will at the 

 first glance be convinced, by the numerous characters 

 they possess in common, notwithstanding the different 

 mode in which they take their food, that they form one 

 connected primary group. This circumstance, therefore, 

 only furnishes a clue for their further subdivision into 

 two secondary groups, separated by distinctions certainly 

 of a lower value than those which separate the Crustacea 

 and Arachnida from Insecta. This is further confirmed 

 by the variations that take place in their mode of feed- 

 ing in their different states ; some from masticators be- 

 coming suctorious (Lepidoptera], and others from being 

 suctorious becoming masticators (Myrmeleon^ Dytiscus, 

 &c.), which shows that this character does not enter the 

 essential idea of the animal. 



iv. Next to Classes and Subclasses we are to consider 

 those groups of insects that are denominated Orders. 

 The characters of these at first were taken principally 

 from the instruments of flight or the absence of them ; 

 and the name appropriated to each Order by Linne, after 



a VOL. III. LETTER XXVIII. 



