30 PEOCEEDINGS OF THE NATIONAL MUSEUM vol. 77 



and Scolosaui'us. Similar ossifications are probably present in the 

 skin of nearly all armored dinosaurs. 



Relationshi'ps. — The genus Palaeoscincus falls readily into the 

 family Nodosauridae as redefined by Lull,-'' and on the basis of the 

 skull structure it would appear to have its closest relationships with 

 Panoplosawus and Edmontonia. These three genera should be 

 grouped to form the subfamily Panoplosaurinae recently proposed 

 by Nopsca,^^ Of the other four genera Dyoplosaurus^ Hierosaurus:, 

 /Scolosaurus, and jStegopelta, included by Nopsca in this subfamily. 

 Dyoplosaurus^ as shown by the recently discovered skull described 

 later in this paper, should now be removed to the Ankylosaurinae, 

 The skull structure of the three remaining genera is unknown and 

 their precise systematic position is much in doubt. The presence in 

 all of transverse rows of nuchal plates is one reason for their provi- 

 sional retention in the Panoplosaurinae. 



The definition of this subfamily may now be emended by the 

 inclusion of the following skull characters : /Skull covered superiorly 

 hy a few large plates; orbits placed far posteriorly; lateral temporal 

 fenestra opening laterally; nostrils later o-ternrdnal ; large quadrate + 

 quadratojugal dermal plate absent. 



While it is not my intention to attempt a revision of the Nodosau- 

 ridae, yet the study of the specimens in hand has disclosed relation- 

 ships of certain American members of this family which appear 

 worthy of record. 



Dyoplosaurus^ as mentioned above, is here transferred from the 

 Panoplosaurinae to the Ankylosaurinae. In establishing this sub- 

 family Nopsca referred to it the following genera : Ankylosaurus.^ 

 Euoplocephalus^ HoplitosoMrus^ Palaeoscincus^ Polacanthus^ Pola- 

 canthoides^ Sarcolestes. With this grouping I can agree only in 

 part. AnJcylosaums^ EuoplocephaJus^ Dyoplosaurus and Anodonto- 

 saurus are true members of this subfamily. On the basis of the skull 

 structure I would amend Nopsca's definition of the Ankylosaurinae 

 by the inclusion of the following cranial characters : Skull covered 

 superiorly hy numerous small plates; orbits placed submedially ; 

 lateral temporal fenestra covered by dermal bone; nostrils terminal; 

 large quadrate -\r quadratojugal dermal plate present. 



The other genera can not be definitely assigned on the basis of 

 their cranial structure, the skull being unknown. Polacanthus., with 

 its complete bony carapace over the sacral region and tail, com- 

 pressed dermal elements unlike any found with the more typical 

 Ankylosaurinae, will, I believe, eventually be found to belong to 

 a distinct subfamily. The American Hoplitosauims is apparently 

 closely allied and should be grouped with Polacanthus. 



=" Dinosaurierrestl, Geologica Hungarica, Budapest, 1929, pt. 5, p. 70. 

 =» Amer. Journ. Sci., vol. 1, February, 1921, p. 123. 



