August 30, 1912] 



SCIENCE 



287 



NOTE ON THE DINOSAUR-TURTLE ANALOGY 



In my paper, " Notes on the Armored Dino- 

 sauria," * I first gave general form to the idea 

 that there is a distinct structural parallel be- 

 tween the armor of turtles and Dinosaurs, but 

 that while in the former fixity and regularity 

 of pattern were early developed, in the latter 

 bizarre patterns were assumed. The main 

 thought was also mentioned in a brief earlier 

 paper of March, 1909, in the same Journal. 



But this view can be made much clearer 

 now. Both Hay and later Von Huene^ agree 

 in pointing out that the so-called parietal ex- 

 tensions of Triceratops are in reality projec- 

 tions formed by the fusion of elements which 

 should be called dermo-parietals. I go further 

 and state that this is not only correct, but 

 that in all probability there are two additional 

 lateral dermal elements fused with the squa- 

 mosals, hypothetically the dermo-squamosals; 

 and the hypothesist is clearly at liberty to go 

 on and say that the horns as well may include 

 equivalent dermo-cornutal elements. Now 

 any such dermo-comutal region may, and 

 both the dermo-parietal and dermo-squamosal 

 region must belong to the same deep dermo- 

 gene layer in the Ceratopsidse that in Pola- 

 canthus and all the Nodosauridse gives rise to 

 what I descriptively call the lumbar-hip-cara- 

 pace which clearly results from the fusion of 

 dermo-iliac plates. For in all these instances, 

 whether in the skull region of the Ceratopsids, 

 or the post-dorsal region of Polacanthus, we 

 see the bones of this primary deep dermal 

 layer undergoing direct fusion with the endo- 

 skeleton, just as in the dorsum of the turtles. 

 Moreover, just as I proved in the case of the 

 early turtles like Archelon and Toxochelys 

 Bauri, the bones of this deep layer bear or are 

 ridden by those of an outer superficial layer. 

 The demonstration of this superficial layer 

 and its run in heels of course explained the 

 origin of the osteodermal carapace of Dermo- 

 cJielys. 



Homologously the outer osteodermal layer is 

 represented in the Ceratopsids by the epoc- 

 cipitals, as Marsh called them, which ride the 



'■American Journal of Science, February, 1911. 



'Neues JaJirbueh, Jahrg. 1911, p. 146, 1912. 



dermo-parietal and dermo-squamosal region 

 and by the similar usually keeled series -of 

 various Dinosaurs. Only in Stegosaurus is it 

 difficult to state whether the two huge rows of 

 dorsal spines belong to the outer, or to the 

 nether dermogene armor-producing layer; 

 while it is not absurd to suggest that the dor- 

 sal plates could possibly result from the fusion 

 of elements of both layers.' Though we 

 should not lose sight of the alternative expla- 

 nation that the skull plates of Anhylosaurus, 

 the horns and frill of Triceratops, the dor- 

 sal plates of Stegosaurus and the dermo- 

 iliac elements of Polacanthus, Nodosaurus 

 and Siegopelta may all be homologous ele- 

 ments of a dominant midline armor arising 

 from the deep dermogene layer and thus in 

 part analogous to the pleuralia of turtles. 



Similarly, going much further afield, it is 

 entirely possible that in the origin of the ex- 

 traordinary supra-occipital crest of Pierano- 

 don brought to light in the course of the ad- 

 mirable studies of Eaton, some strictly dermal 

 element has played a part. And, indeed, re- 

 sponsive or counter growth of the endoskele- 

 ton finally resulting in fusion with dermo- 

 gene elements and the complete obliteration 

 of sutural lines, is of common observation in 

 the Vertebrata, being essentially a senile 

 course of development, which has to do with 

 the aging of races quite as much as the pro- 

 duction of purely protective features. The 

 studies, of Beecher on the origin of spines have 

 interest in this connection. 



Going back to the first premise: As noted, 

 in strong contrast to crocodile-like reptilian 

 types with an outer dermogene bone-producing 

 layer only, the turtles originally had promi- 

 nently developed, both the outer and nether 

 dermogene layers. But they early tended to 

 strengthen and use the under layer only along 

 very conservative lines, and in their history 

 never developed cranial armature, save in the 

 comparatively recent and altogether aberrant 

 Meiolania. On the contrary, as fusion of the 

 dermogene with the chondrogene elements of 

 the carapace and plastron went on, the armor- 

 less head became more or less retractile; 

 while the carapace and plastron, though of 



