NO. II LEGS OF PRIMITIVE ARTHROPODS EWING I9 



end there is a very strong, transverse apodeme on the outside margin. 

 The apodeme does not extend entirely across as an unbroken ridge, 

 but in it there is a gap. At this gap there is situated a prominent 

 fusiform appendage, the much discussed " coxal appendage." The 

 second segment is small and much curved. It is hinged toi the first 

 Iw a rocking double hinge. It extends upward and outw^ard. The 

 third segment is large and long and is immovably attached to the 

 second. Distally it supports the fourth by means of a double hinge. 

 The fourth is about as long as the third but is smaller. Beyond the 

 fourth segment are three small ones that move together as one. The 

 distal one of these bears the two equal claws. An examination of the 

 attachments of the muscles shows that the fourth segment gives origin 

 to the last muscle of the leg, the flexor of the claws. It should, 

 therefore, be regarded as the tibia. Then the three small ones distal 

 to it should be considered as the tarsus and the large one proximal 

 the femur. Because the small second segment is immovably fused 

 with the femur, gives origin to muscle fibers that flex the tibia, and is 

 moved by a double hinge on the first segment, it must be considered 

 as the trochanter. The basal segment is regarded as the true coxa. 

 Although it is very large its musculature and the attachment of the 

 appendage to it indicate clearly its homology. The coxa itself is 

 hinged to an incomplete segment which has a stout apodeme. This 

 incomplete segment gives rise to levator muscles of the coxa. It 

 should be regarded as a functional subcoxa. 



In the Lepismatidae, Thermobia is taken as the type. In this genus 

 the leg (pi. 5, fig. 14) is similar to that of Machilis, but lacks first 

 of all the appendage to the coxa. The subcoxa here is a complete 

 segment, and frequently shows two psevidosegmental sutures on its 

 antero-dorsal aspect. The coxa lacks not only the appendage of 

 Machilis but also the conspicuous transverse apodeme. The trochanter 

 differs from that oi' Machilis in two ways; it frequently gives rise 

 to a very short muscle going to the base of the femur as shown by 

 Escherich (1905) or again it may be entirely without muscles, even 

 without fibers from the flexor of the tibia. Such a trochanter has 

 been drawn by Carpenter (1916). The femur, tibia, and tarsus are 

 essentially the same as in Machilis. 



The legs in Campodeidae and Japygidae are essentially the same 

 in segmentation and musculature. Those of Campodea are taken as 

 the type (pi. 6, fig. 15). The Campodea legs are very slender as 

 compared with those of Machilis and have the following outstanding 

 differences : The coxa is greatly reduced, very short and is without 

 the transverse apodeme and appendage ; the trochanter is a free- 



