NO. I INSECT THORAX SNODGRASS 5 



basic theory is redistilled in a new form. The product, it is hoped, 

 when tested, will be found to be a closer approximation to the truth 

 than any of the ingredients, but it will fail of its purpose if it does 

 not act as a stimulus to the further study of the facts bearing on the 

 evolution of the insect thorax. 



PRIMARY SEGMENTATION 



In the study of insect evolution, we go back to a soft-skinned, 

 worm-like creature with the body divided into a series of cylindrical 

 parts, or segments, each of which bears a pair of lateral or ventro- 

 lateral appendages. In an animal of this sort the intersegmental rings 

 are the lines of attachment of the principal longitudinal muscles; in 

 fact, the circular grooves separating the segments, and, therefore, the 

 segments themselves are determined by the muscle attachments — in 

 other words, the body segmentation corresponds with the muscle seg- 

 mentation. If, at an earlier phylogenetic stage, the animal was unseg- 

 mented, it would seem that the body segments, or somites, must have 

 resulted from the division of the muscle layer into muscle segments, 

 or myotomes. 



Starting with the insect's ancestor as a soft-skinned, segmented 

 animal, resembling in its segmentation the soft-bodied larvae of some 

 modern insects, we must believe that its body segments then corre- 

 sponded with its muscle divisions (fig. 2 A). This type of body 

 division we may call primary segmentation. The muscles, extending 

 between the intersegmental rings, are segmental in arrangment. Al- 

 ready, we assume, the creature possesses well-established dorsiven- 

 trality and cephalization ; i. e., one surface, the back or dorsum, is 

 normally uppermost, and the opposite surface, the venter, is down- 

 ward, while, in the usual progression, one end, the head end, is 

 forward. 



SECONDARY SEGMENTATION 



The modern adult arthropods, unlike their hypothetical ancestors, 

 are in general hard-shelled animals ; they have developed an external 

 skeleton formed of calcareous or chitinous matter or both deposited 

 in the ectoderm, and the hardening of the body wall has had a far- 

 reaching consequence on the structure of the segments and on the 

 general mechanism of the animal. The skeletal deposits have taken 

 the form of segmental plates (fig. 2 B), the principal plates in each 

 segment being a dorsal one, or terguni (T) and a ventral one, or 

 sterntitn (S). These two plates are separated on the sides of the 

 segment by a membranous pleural area. The hardened parts of the 



