OF DIDELPIIYS VIRGINIANA. 75 



curled around the fore paws. Considerable twisting, or rotation of the vertebrae upon 

 each other, can be borne without inconvenience. The perfection of the "fifth hand" is 

 brouo-ht to mind as forcibly by its bones, as by the admirably adapted sot of muscles, 

 the closely packed bundles of tendons, and their dense fascial and cutaneous envelopes. 



The first caudal is nearl}^ a repetition of the last sacral, except as to its transverse 

 processes. These resume the thin, expanded, laminar character of those of the lumbars ; 

 but increase in lateral projection or length, and are horizontal, not oblique. The spinous 

 process sufiers marked diminution both in height and length ; the pre-zygapophyses are 

 anchvlosed ; the posterior ones, though small, bear perfect articulating facets. The under 

 surface of the centrum is concave lengthwise. 



The next four (second — fifth) caudals are similar to each other, and have the following 

 characteristics : The spinous process becomes rudimentary with the first of them (second 

 coccygeal), and disappears with the last (fifth coccygeal). The transverse are broad, long, 

 thin plates, of a squarish shape, slightly constricted at the base ; they call to mind the 

 corresponding process of the atlas ; their longitudinal extent or width increases, from 

 before backward, coincidentl}^ with a decrease of lateral projection ; that of the second coc- 

 cygeal is longest and narrowest. The anterior articukr are long, conical, flaring processes, 

 completely embracing the post-zygapophyses ; both bear perfect f icets. The neural arches 

 are very short, and already have greatly diminished the calibre of the neural canal. The 

 centrums are long for their breadth; the disparity increases from before backward with 

 successive vertebrae, at the same time that their under surfaces become more and more 

 flattened. 



The degradation of the vertebrae, already apparent, is suddenly increased, and becomes 

 very obvious, by the modification of the sixth caudal, which is the first to decidedly assume 

 the physical characters that are to characterize the rest. Its transverse process changes to 

 a thin lamina, running the whole length of the bone, wider anteriorly than behind ; the body 

 itself of the vertebra appears as if it were merely a longitudinal thickening of this lamina 

 along the median line. Anterior articulating processes are perfect, as hitherto, in tlie cau- 

 dal series ; but post-Z3^gapophyses have become rudimentary, and articulation with the 

 seventh is effected only by the centrum. The rudimentary neural spine is only indicated 

 by a thin, low crest along the middle line. Neurapophyses, however, have not entirely sub- 

 sided, so that this vertebra is still perforate ; but the canal barely admits the passage of a 

 bristle, and appears not to transmit a continuation of the neural axis. The neural canal of 

 the fifth terminates at the distal extremity of the bone in a somewhat difterent manner 

 from that observed in antecedent vertebrae; the neural arch, leaving the articulating 

 processes, extending upward, and flaring outward, is separated by a deep fissure from the 

 centrum along the posterior third of the vertebra, so that the canal has a lateral as well as 

 a terminal opening. The seventh caudal is impervious, and in other respects completes 

 the changes that began with the sixth ; and successive ones are no longer, teleologically, 

 vertebrae, but simply the phalanges of the opossum's fifth hand. 



The remaining caudals resemble phalangeal internodes, both in appearance and function. 

 The next succeeding half dozen are longer, both absolutely and relatively to their width, 

 than any of the preceding ones; but they gradually shorten, and more rapidly become 

 attenuated. The transverse processes appear as a ridge on either side of the seventh — 

 tenth, suffering constriction at the middle, so that the bones, viewed from above, present an 



