VERTEBRAL NUMBERS IN SHARKS — SPRINGER AND GARRICK 77 



posterior monospondylous centra and the anterior diplospondylous 

 centra is similarly variable though reasonably constant for any one 

 species. 



Reporting these relative centrum lengths is obviously of value but 

 offers difficulties. Because many of the specimens we used were 

 curled or distorted, radiogi-aphs of them show oblique images of 

 some centra. The degree of obliqueness may vary from one region 

 of the vertebral column to another; we, therefore, see the likelihood 

 of introducing considerable error by giving dimensions of centra that 

 are not adjacent to, or at least very close to, each other on the radio- 

 graph. For this reason, we do not report the length of anterior 

 relative to posterior monospondylous centra. To compare the length 

 of a posterior monospondylous centrum with that of an anterior 

 diplospondylous centrum (thereby indicating the relative prominence 

 of the transition point between monospondyly and diplospondyly), 

 we measured the length of the penultimate monospondylous centrum 

 and the first diplospondylous centrum and then obtained a ratio, 

 listed as "A" in the results and tables, by dividing the length of the 

 former by the length of the latter and multiplying by 100. We 

 chose the penultimate monospondylous centrum because it is usually 

 more representative of the posterior monospondylous centra (the 

 ultimate monospondylous centrum is often reduced in length even 

 though noticeably longer than the first diplospondylous centrum). 

 The first diplospondylous centrum was chosen because there is often 

 a regular alternation of centrum length in the diplospondylous region. 

 In deciding the transition point from monospondyly to diplospondyly 

 when differences in length of the centra bordered on being imper- 

 ceptible, we may be in error occasionally though we think that any 

 error thus introduced is slight. 



To give an indication of centrum shape, we have divided the length 

 of the penultimate monospondylous centrum by its diameter (appear- 

 ing as height or width on the radiograph) and have multipled this 

 ratio by 100 to yield the values listed as "B" in the results and tables. 



Results 



Our data on shark vertebrae are reported in table 1. Table 2 

 condenses these data so that the subordinal, familial, and generic 

 limits, but not those of the individual species, can be more easily 

 visualized. 



We have representative samplings of six of the seven suborders of 

 sharks and a partial sampling of the seventh. None of these sub- 

 orders can be distinguished on its vertebral counts alone. In terms 

 of total number of vertebrae we found a range of from 60 to about 419. 

 The lowest total count is for Squaliolus laticaudus and the highest for 



