96 



SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 8o 



knee joint to which the flexor muscle is here often attached. As 

 already pointed out, however, this interpretation seems entirely un- 

 necessary, since there is no evidence that the tibia and tarsus are 

 not primitive limb segments, as the tarsal muscles indicate. In coleop- 

 teran larvae with a five-segmented leg (fig. 43 B), therefore, it is 

 most probable that the tibia and tarsus are united, as also is indicated 

 by the attachment of the first branch of the pretarsal muscle {X). 

 Muscle insertions in many places are upon small chitinizations of an 

 articular membrane, rather than directly upon the part to be moved 

 by the muscle. 



The idea that the pretarsus of the insect represents a primitive leg 

 segment has not been generally accepted, some entomologists regard- 

 ing it as a sixth segment of the tarsus, while others see in the terminal 



k X 



nar 



Fig. 44.— Terminal foot structures of Tliysamira. 



A, hind claws and end of tarsus of Jal^yx, dorsal view, showing rndinientary 

 median claw (/>). B, same, ventral view, showing large unguitractor plate 

 (Utr). C, foot of Lcpisma, lateral, with dactylopodite-like median claw {Dae) 

 to which is attached tendon (.r) of depressor muscle. D, same, end view. 

 E, median claw, or dactylopodite, of Lcpisma, with piece of depressor tendon (jr). 



structures only special developments of the last tarsal segment. The 

 present writer, however, would now adopt the view so well stated 

 and so fully illustrated by de Meijere (1901), that the pretarsus, 

 including the claws, the arolium, and all accessory parts of the foot 

 is a development of the dactylopodite of the generalized arthropod 

 limb. 



A pretarsus having the form of a simple claw-like dactylopodite 

 occurs in the Protura, in some of the Collembola, in caterpillars, 

 sawfly larvae, and in most beetle larvae. In the Lepismidae each pre- 

 tarsus has three claws, there being a slender decurved median claw 

 between the two lateral claws (fig. 44 C, D, Dae). The median claw 

 is a hollow structure (E) with the unguitractor tendon (.r) attached 

 to its base, and, if the lateral claws were lacking, it would constitute 

 a miniature but typical dactylopodite. In Japyx the ventral part of 

 the pretarsus apparently has become developed into a large unguitrac- 

 tor plate (fig. 44 B, Utr), while a point that is perhaps its tip remains 



