EDENTATA OF THE SANTA CRUZ BEDS. 155 



armadillo-like than glyptodont. Another and even more obvious differ- 

 ence from the preceding genera is in the shape of the tail as a whole, 

 which is very slender in the hinder portion and tapers gradually to a 

 point, instead of being thick, club-like and truncated at the end. This 

 specimen cannot be referred to any of the preceding genera and it is 

 assigned to Asterostemma, because in the Ameghino collection is a frag- 

 ment of a similar tail-sheath, which was found associated with carapace- 

 plates of the type originally referred to this genus. 



In the same collection is the facial portion of a skull, also associated 

 with similar plates, which displays a number of significant peculiarities. 

 One of these peculiarities is in the form of the upper teeth (Plate XXIV, 

 fig. 7) of which only * has a trace of the division of the postero-external 

 lobe into two pillars, while on the succeeding teeth ( & & ) there is no sign 

 of such a division. Asterostemma is the only known Santa Cruz glypto- 

 dont of which this is true and the significance of the fact lies in the resem- 

 blance of these teeth to those of the later representatives of the order, in 

 which the division does not occur, save in * of Sclerocalyptus. 



Concerning the skull, Ameghino remarks that "the premaxillaries are 

 rudimentary, not reaching the nasals" ('98, 210). With this description 

 I am not able entirely to agree. I "think Dr. Ameghino has been deceived 

 by the fact that the premaxillaries have been lost from the specimen in 

 question, leaving a smooth edge of the maxillaries, which thus appears to 

 form the true boundary of the anterior nares. On the other hand, a nasal 

 process of the maxillaries seems to have prevented any contact between 

 the premaxillae and the nasals. The palatine processes of the maxillae are 

 deeply notched by the incisive foramina, which extend as far back as the 

 middle of -, and are thus very much larger than in any of the preceding 

 genera. The nasals are longer, narrower, more pointed anteriorly, and 

 extend farther in advance of the maxillaries than in Propalczohoplophorus. 

 Altogether, this region of the skull is quite characteristically different from 

 the muzzle in the other contemporary genera. 



While most of the Santa Cruz glyptodonts appear to be without phylo- 

 genetic importance and probably were not directly ancestral to any of the 

 later types, Asterostemma may prove to be the common ancestor of many 

 of the Pampean genera. The tail-sheath is the least specialized yet dis- 

 covered in the order and the character of the teeth is very suggestive, 

 but the materials are still insufficient to justify a more positive statement. 



