1877.] 203 [Garman. 



is to be expected, contains the greatest variety of pelvic modifications. 

 Through all of the second the ventrals seem to retain the same uses 

 and about the same position, with respect to the axis of the body, 

 consequently the variations in shape of the pelvis attending the dif- 

 ferent uses of these fins in the first group do not appear. Instead, it 

 is such as are due to difference in the general build of the body that 

 become prominent in this section. With the lines indicating the 

 direction of use of the ventrals at right angles to that of the pecto- 

 rals, and parallel to the vertebral axis in nearly or quite all of the 

 species of the division, there are no apparent necessities for variation 

 in the position of the articulations other than result from the changes 

 of curvature of the shaft. In the first section the processes for artic- 

 ulations vary in position from the hinder side to the end of a shaft 

 nearly straight as the fins are used more toward the side than back- 

 ward. In the second they vary in position toward the end as the 

 curvature of the shaft places the ends more directly backward. 



The pelvis of Rhinobatus resembles that of Raja; it is nearly 

 straight, strong, depressed, and has short stout processes from the 

 ends on each side of the viscera. 



Torpedo presents an extreme case of the backward arch, as well as 

 of the side attachment of the ventrals. The cartilage extending 

 forward on each side of the abdomen seems to be a direct continua- 

 tion of the shaft (pubes and ischia), though, if we consider itsbegin- 

 ning to be at the articulations, it corresponds to the ilium. 



Narcine, of the same family, has a shaft more straight but still 

 concave anteriorly ; the ilium is represented by long slender for- 

 ward extensions between the wall of the body and the abdominal 

 cavity, and also by a similar extension posteriorly, — an iliac devel- 

 opment in a measure resembling that of the bird. 



Raja has a strong straight depressed shaft, with processes in front 

 of each extremity. This form is common to the entire family. The 

 greatest variation occurs in Sympterygia, and a peculiar type (ilia- 

 lacorliina mira, described below) from the Mejillones, of which but a 

 single specimen has come to my notice. 



Sympterygia (see S. acuta below) has a slight forward curve and 

 long slender anterior iliac extensions, somewhat resembling those of 

 Narcine. The specimen from the Mejillones, while agreeing with 

 Sympterygia in respect to absence of rostral cartilage, bears in other 

 respects a closer resemblance to Raja. It has a straight shaft, and 

 lateral prepelvic cartilages intermediate in size between those of the 



