1.50 NATURAL ARRANGEMENT OF INSECTS. 



ami pointed." Again, in reference to the Coccus Ado- 

 nidiim, he remarks, "It has the appearance of a small 

 millepede, or Oniscus; being of an oval shape, and slightly 

 convex above, with the body divided into many trans- 

 verse segments, projecting sharply on the sides, and 

 furnished with small processes or points." Now, these 

 words, intended to describe a Coccus, are equally just 

 when applied to an Oniscus; and, with very little vari- 

 ation, would be suited to the armadillos among qua- 

 drupeds, and to the larvae of the Ericinian butterflies. 



CHAP. IV. 



THE HYMENOPTERA. 



(140.) The order Hymenoptera, according to the 

 sei-ies laid down in our last chapter, follows that of the 

 Hemiptera. If a zoologist, versed in the other classes 

 of animals, was called upon to select that character which 

 he considered as the most essential, it would be, that 

 these insects have an appendage to the tail, either in 

 the form of a sting, or an oviduct. This character, in 

 fact, is so general, that it only disappears in a portion of 

 the order (Tenthredo Linn.) which is aberrant, and 

 where we should naturally expect some one or more of 

 the typical characters would be lost. It is by this, also, 

 that the analogy of the Hymenoptera to the thysanuri- 

 form larva is produced ; and both these, as will be clearly 

 shown, are modifications of the ungulated type of qua- 

 drupeds. Other characters, not so universal as this, are 

 exhibited in their wings and mouth. The first, where 

 they exist, are membranaceous and transparent, furnished 

 with corneous veins, arranged longitudinally and trans- 

 versely, leaving large spaces, — a structure very different 



\ 



