54 



PEOr. E. W. CLATPOLE ON PTEPvASPIDIAN FISH IX 



above them in the upper division of the vascular layer, and below 

 the minute channels of the upper and outer surface, with which 

 they are apparently continuous. These highest canals, to which 

 possibly the term furrows would be more applicable, have no 

 branches or outlets, so far as I have been able to determine, except 

 the passages just mentioned, nearly as wide as themselves, whereby 

 communication is effected with the vessels lying below them, and 

 already described. Both may be readily seen sectionaUy and longi- 

 tudinally by cutting thin slices of the shield in the proper direction 

 (see figs. 2, 3, 4). 



■Sections of the Shield of Palseaspis. 

 Kg. 2. 





I^ig. 3. 



Fig. 2. Vertical longitudinal section showing (a) basal layer ; {b) cellular layer ; 



(c) vascular layer, {}) lower portion, (^) upper or tubular portion. 

 Fig. 3. Vertical transverse section, showing (a) basal layer ; (6) cellular layer ; 



(c) vascular layer, Q-) lower, (^) upper portion. 

 Fig. 4. Horizontal section through vascular layer of shield, showing the lower 



branching vessels, the upper vessels or channels, and the tubules (seen 



from below). 



Lastly, there yet remains to be described the most remarkable 

 part of this outer or upper vascular layer. To this I have given 

 the name of the tubular division. From the larger and lower of the 

 two sets of vessels just described and from its upper side there issue 

 a vast number of minute tubules which branch and run to right and 

 left, and upward from the parent vessel, forming a minute and beau- 

 tiful forest-like growth. Pilled as they are with opaque matter 

 they stand out distinctly, like dendrites on marble, on the light yel- 

 low surface of the material of the vascular layer. They do not form 

 any connexion with the large channels or furrows lying to right 



