Insofar as the lengths of the legs show, _Dulus_ and _Phainoptila_ are the most generalized of the Bombycillidae, since the relative length of leg is approximately the same as that of more generalized birds such as warblers, crows and thrushes of similar locomotory habits. In other words, _Dulus_ and _Phainoptila_ have remained unspecialized, in contrast to the waxwings in which adaptive changes fitting them for a perching habit have taken place. _Ptilogonys_ and _Phainopepla_ are intermediate in length of leg between _Phainoptila_ and _Bombycilla_, and _Ptilogonys_ and _Phainopepla_ have progressed from life on the ground toward the perching habit. _Bombycilla cedrorum_ is more specialized than is _B. garrula_ in shortness of leg, and the reduction is comparable, as is noted above, to that in the legs of _Tachycineta_. In birds which have the legs much modified for walking or for hopping in the brush, such as _Polioptila_ and _Eremophila_, it is noteworthy that the distal segment, the tarsometatarsus, is the longest, whereas in birds such as _Myiarchus_ and _Tachycineta_, that do not utilize the limbs in this manner, the tibiotarsus, the middle segment, is the longest. Mammals much modified for walking or hopping likewise have the proximal segment, the femur, short, and the distal segment long (Howell, 1944). The waxwings have all of the segments short; these birds are modified for strong and sustained flight. Their hind limbs are used principally for landing devices and for perching. No one element of the leg has been shortened much, if any, more than any other. [Illustration: Fig. 44. Graph showing relative lengths of bones of the leg. The percentage values are shown on the axis of the ordinates. A. _Bombycilla cedrorum_; B. _Bombycilla garrula_; C. _Dulus dominicus_; D. _Phainoptila melanoxantha_; E. _Phainopepla nitens_; F. _Ptilogonys cinereus_; G. _Ptilogonys caudatus_. a. femur; b. tibiotarsus; c. tarsometatarsus; d. total.] _Arm-trunk Percentages._--Tables 1 and 2 show the total length of the arm, and lengths of the separate arm elements, relative to the trunk. Table 3 gives the corresponding lengths for birds other than the Bombycillidae. Total length of arm was obtained by adding together the lengths of the humerus, ulna, and manus, and by dividing the figure thus obtained by the length of the trunk as was done for leg lengths in tables 4 and 5. The method of adding together the component parts does not give the entire length of the wing, since the length of the feathers, which add effectively to the total length, as well as do the lengths of the small carpal elements, is lacking. [Illustration: Figs. 45-46. Outlines of wings. x 1/2 45. _Ptilogonys caudatus_, showing relation of outline of wing to bones of arm. 46. _Bombycilla cedrorum_, showing relation of outline of wing to bones of arm.]