154 University of California Puhlications in Zoology [Vol. 13 



In side view the complexity of the dorsal scale becomes apparent. 

 The body of the scale is intimately related to the base, the neck or 

 pedicle (ped.) beinpf here much thicker and less attenuated than that 

 of the ventral type. 



The termini of the arms of the cross appear as enlarged projections 

 or nodules. The posterior arm {post, a., fis:. D), which is characteristic 

 of the dorsals and appears but rarely in the ventrals (see central 

 scale, pi. 8, fig. 2), is here seen to especial advantage. Evident in such 

 a view are the crests of the scale. On the body of the scale the ter- 

 minal branches of the primary and secondary crests are seen to extend 

 to crests from the base. The crests on the base and neck we may 

 designate as columns (col.) which extend from the termini of the 

 arms do\^Ti the neck to the periphery. From this view it is further 

 to be noted that on the posterior arm terminal crests are also to be 

 found and that these are the remnants of a posterior crest w^hich 

 extends to the tip of the spine. 



From this it is clear that in Heterodontus the enlarged columns 

 of the base, although they may (or may not) be continuous with the 

 crests of the arms, are in no wise the same structures as they are in 

 some of the elasmobranchs such as AcantJiias vidgaris and Centro- 

 phorus sp. (see Steinhard, 1903, pi. 1, fig. 2'5, and pi. 2, fig. 38). 



Modifications in Form of the Dorsal Type of Scale. — The scales 

 making up the supraorbital crest show profound modification in form 

 when compared with those of the dorsal type (pi. 9, fig. 7). To the 

 unaided eye they appear as large, round nodules which may attain a 

 diameter considerably in excess of that of a normal scale. A closer 

 examination reveals the fact that this shape is due to the enlargement 

 of the whole body of the scale and of the spinule located upon it. In 

 some of the smaller of these the spine may arise abruptly from the 

 central part of the body, the four crests converging at the tip of the 

 spine. In others the spine, more or less blunted, slopes gradually 

 posteriorly, and in still others the spine is relatively of immense size, 

 so large, in fact, that its detail can be made out by the unaided eye. 

 In such a scale the spine often leans to the right or to the left. Cor- 

 related with this heavier spine go similar modifications of the arms. 

 The simplest of these modifications is seen in the bifurcation of the 

 anterior arm, thus adding a fifth. In others an unpaired anterior arm 

 may arise secondarily, making in all six arms. 



Between the average and the more highly modified supraorbital 

 scales various degrees of asymmetry obtain. Superficially, this is first 



