682 JOURNAL, BOMBA Y NA TURAL HISTORY SOCIETY, Vol. XI. 



to continue along the corresponding surfaces of the two branches ; but 

 absolute continuity is seen only in the case of the left branch (fig. 3). 

 In this laterally compressed structure the ventral series extends further 

 up the left side than the right, so that only one series of small scales 

 is present on this side, while two series separate the ventral and dorsal 

 large scales on the right side. The scaling of the dorso-ventrally 

 compressed right branch is very irregular at the forking, but at "6 cm. 

 therefrom a distinct series of broad scales commences to run along 

 the ventral surface (fig. 4). These connect on the right side with a 

 dorsal series of large scales, which perhaps may be regarded as a con- 

 tinuation of the dorsal series of the undivided proximal portion of the 

 tail. Two series of small scales complete the scaling by covering the 

 left dorsal surface. 



Branches of Specimen B. — The dorsal series of large scales of the 

 undivided proximal portion of the tail is continued along the dorsal 

 side of the dorsal branch (fig. 5), but twists towards the left side as it 

 proceeds. The ventral series of large scales of the undivided proximal 

 portion is continued along the ventral side of the ventral branch (fig, 6) 5 

 but twists towards the left side as it proceeds. The dorsal branch has 

 thus a single series of large scales on its left side and from three to 

 two series of small scales on its right side. The ventral branch has the 

 left-central series of large scales above mentioned, while on its sides 

 two series of small scales separate the large scales from a right-dorsal 

 series of large scales which starts from the irregular scales at the forking. 

 Whether this last-named series can be regarded as having anything 

 to do with the dorsal series of the undivided portion of the tail is 

 uncertain. The starting of a series of large scales from small 

 irregular scales at the forking resembles the condition already describ- 

 ed for the right branch of specimen A. The interest of the matter is 

 that in both specimens both branches of the tail exhibit a tendency 

 towards the establishment of a symmetrical scaling regardless of the 

 plane in which the forking occurs. In other words, though the forking 

 is horizontal in one specimen and vertical in the other, the individual 

 branches very soon assume an arrangement attaining or approximating 

 to dorsal and ventral single rows of large scales while the sides are 

 covered by small scales. The twisting of the series of scales round 

 the long axis which is seen in the present examples recalls a similar 



