78 



University of California. 



[Vol. 3 



that it has been necessary first to discover whether their position 

 in the matrix can furnish a clew to their normal place in the arch. 

 As the skeleton is resting on its left side, it is probable that 

 the arch and limb bones on that side were next the sea bottom 

 and immediately covered by the body, and would be subject 

 to less disturbance than the corresponding elements of the upper 

 or right side. The latter would be separated by a large mass of 

 flesh from the sea floor, and in settling down after the body had 

 decomposed might easily have been separated or moved out of their 

 natural order. As the arch bones lie on the slab, one of them, 

 Isl, is covered by a femur, another, PI, seems to have had ribs 

 lapping over it, and a third, II, is partly covered by Isr. These 

 three bones seem to have been on the lower side of the body. 

 They are, moreover, grouped about the head of the femur Fl in 

 such a way as they might be if they belonged with it. The other 

 three bones are grouped with the corresponding ends near each 

 other. Of these elements PI and Pr are anterior to Isl and 

 Isr. II is nearer the vertebral column than the others from that 

 side, and is probably the ilium. Ir does not occupy the same 

 relative position, but may be displaced. PI and Pr may be con- 

 sidered tentatively as the pubic bones, Isl and Isr as the ischia. 



In comparing these elements with those of other swimming 

 saurians, the supposed pubes and ischia are found to show some 

 resemblance to those of Plesiosaurus and Cimoliosaurus and also 

 to Palaeohatteria , though their positions are reversed. The 

 elements which have been considered as the ilia are unlike any 

 bone of the Ichthyosaurian or Plesiosaurian pelvis that has been 

 accessible for comparison. An approach in form would be found 

 in an inverted pubis of Ichthyosaurus, the ilium of the later 

 Plesiosauria, or in the most common form of ischium of the 

 Pythonomorpha. From all of these it differs so much that the 

 writer has been unable to do better than depend for tentative 

 determination upon their present position in the matrix, as previ- 

 ously stated. As careful a preparation as can be made of the 

 proximal ends of the pelvic elements of perrini fails to show 

 definitely more than a single articular surface, excepting in the 

 ilium; so that it has not been possible to ascertain how they 

 could be fitted together to receive the head of the femur. In 



