22 THE PERIODIC GROWTH OF SCALES IN GADID.^ 



cavity. In the fifth degree, the spines have the same external form as in the 

 last case ; but they are not composed of homogeneous tissue similar to that of 

 the scale, but of dentine, in which canaliculi extend from the central canal to 

 near the surface. Such a structure is found, for example, in the spine of 

 Hypostoma. As to the dimensions of spines on scales and their growth, he 

 says that in passing from tlie free border to the centre of the scale they 

 gradually lose their volume, but in a transitional manner. The dimensions of 

 spines increase with the age of the fish in a marked degree. The number 

 of spines also varies in different regions of the body and with age. By a com- 

 parison of scales from the same fish one finds that the number of spines varies 

 only slightly in points from the same or adjoining regions of the body ; but 

 those scales from different regions show considerable variations as to the 

 number of spines. There are, however, exceptions to this rule (dab). The 

 number of spines as of concentric ridges is usually greatest in scales from the 

 median region of the side. 



In those regions in Avliich scales tend to be rudimentary, they also tend to 

 lose their spines, and thus become cycloid. The fact seems almost certain, 

 that there does not appear to be a single ctenoid fish in which one would not 

 meet cycloid scales on certain points of its body. Baudelot brings forward 

 some facts to show that new spines form themselves behind those already 

 existing on the posterior border. The spines and concentric ridges are homo- 

 logous productions, and growth of both takes place in the same direction. 

 According to Baudelot, then, spines are products of the same nature as the 

 concentric ridges ; they are ridges which have become very prominent, and cut 

 into transversely in such a manner as to constitute a series of prolonged 

 spines, each with a distinct base. In support of this hypothesis he brings 

 forward the following facts : — 



In many scales, such as those of the perch and mullet, the edge of the 

 concentric ridges presents a series of very distinct microscopic indentations, 

 and in some ctenoid scales the spines are so small as to represent only stronger 

 indentations of the ridges of the posterior region which have become very 

 prominent. In many cycloid scales, such as those of the carp, the posterior 

 region shows a series of tubercles arranged with as much regularity as spines, 

 and which present the greatest analogy to these structures. These tubercles 

 are, however, only partial thickenings of concentric ridges. In the same fish 

 scales become altered and pass from the ctenoid to the cycloid condition, and 

 in that case it frequently happens that the spines become replaced by simple 

 ridges, a substitution which is a clear proof of the homology of spines and 

 concentric ridges. Among Pleuronectidae, in which some are ctenoid (sole, 

 dab), and others are cycloid (brill, flounder), the scales of cycloid forms 

 frequently show in the posterior area, instead of rows of spines, distinct islets 

 of calcareous matter, each supporting a fragment of concentric ridge. When 

 these islets of calcareous matter become straitened and more regular, they 

 evidently result in spines. 



6. The grooves (sillons) on scales. This term has been given to very 

 narrow grooves or trenches which are supposed to have been excavated at the 

 expense of the superficial layer of the scale. These grooves are not present in 



