1 68 PATAGONIAN EXPEDITIONS I PALEONTOLOGY. 



these same regions they are perforated by a canal, which opens by a single 

 foramen upon the floor of the neural canal and by two large foramina 

 upon the ventral face of the centrum. These foramina vary in size in the 

 different genera and, to some extent, also with age, but are invariably 

 present and conspicuous. The neural spines of all the trunk-vertebrae 

 have a backward inclination, which is much more decided in some genera 

 than in others ; they also vary much in length, being disproportionately 

 short in some of the smaller species, but they are always relatively short 

 and slender and thus form one of the most striking contrasts between the 

 Santa Cruz and the Pleistocene Gravigrada. Another notable difference 

 from the latter is in the development of the accessory zygapophyses on 

 the posterior thoracic and lumbar vertebrae. As has already been shown, 

 these processes are quite as complex in the Santa Cruz armadillos as they 

 are in recent genera, but this is not true of the Gravigrada. In the Meg- 

 alonychidce, so far as they are known, there is only one additional pair of 

 prezygapophyses, which are placed upon projections of the neural arch 

 below the primary pair; the additional postzygapophyses are upon the 

 transverse processes. The Planopsidce, of the Santa Cruz are somewhat 

 further advanced and have added a second accessory pair of pre- and of 

 postzygapophyses in the lumbar region. With the possible exception of 

 the Mylodontidcz, concerning which nothing is known in this respect, all 

 the trunk vertebrae are free. 



The sacrum is short and generally consists of five vertebrae, of which 

 the first three are intimately connected with the ilia and the last one with 

 the ischia; they have broad, depressed centra and a low, wide neural 

 canal ; the neural spines are fused into a continuous ridge, with thickened 

 free border and the zygapophyses are obsolete, except on the first and last. 



The tail is of moderate length, but very heavy, especially in the Plan- 

 opsidce, and tapers rapidly to the end ; probably it served as a support 

 when the animal rose upon the hind-legs, as is believed to have been 

 the case in the great Pleistocene forms. The anterior caudal vertebrae 

 are very large and have very long and heavy transverse processes, but 

 posteriorly the vertebrae diminish rapidly in size and all the processes are 

 reduced. A complete series of well-developed chevron-bones is present. 



(6) The ribs are usually of only moderate breadth and therefore very 

 unlike those of the great Pleistocene genera, but they are elongate and 

 form a capacious thorax, which is quite different in shape from that of 



