Structure o/'Stromatopora. 257 



indicative of the holes of the polyps, being thus reduced to 

 various diameters below 1 -600th inch, causes them to look 

 like apertures in the midst of a layer composed of the puucta 

 (i. e. ends of the rods alone) . Or, in the curvilinear structure, 

 the interspaces representing the ccenosarcal cavities become 

 so uniformly contracted that both the horizontal and vertical 

 sections indicate a composition of vertical tubes alone, in 

 juxtaposition like those of Favosites gothlandicus, but with 

 the stellate venation to be mentioned hereafter and without 

 the tabula:, though still communicating with each other as 

 freely as the ccenosarcal cavities in the more typical form. 



Again, although the differences between the rectilinear and 

 curvilinear structures are so marked, they are very analogous 

 in their growth, inasmuch as the end-to-end anastomosis of 

 the old fibre with the new in both instances leads to a conti- 

 nuity of structure in the mass, differentiated only by the 

 characters above mentioned. 



Lastly, the " vertical rods " may be thick or thin ; and 

 although par excellence confined to the rectilinear (PI. XV. 

 fig. 3, a a), similar ones also occur in the curvilinear structure 

 but not always, and therefore are not inserted in the illustra- 

 tions ; they are, however, just as much in connexion with the 

 curvilinear structure, which they simply pierce, as they are 

 with the triangulated layers of the rectilinear structure, which 

 they as simply unite together. They may also present the 

 same appearance of original hollowness. Mr. Champernowne 

 informs me that he has several specimens bearing these 

 "rods" both with and without the presence of the tubes 

 which characterize the specimens called " Gaunopora j" so 

 that the rods do not depend upon the presence of the tubes. 

 In my own small collection the only specimens of curvi- 

 linear structure (viz. two) which bear the "rods," happen 

 to be accompanied by the tubes of Gaunopora. But the most 

 remarkable feature in them is that their axial structure, as 

 well as the connecting fibre being transparent, indicates that 

 they were originally hollow (figs. 2 and 3, eee). 



How far this hollowness was in connexion with the cceno- 

 sarcal cavities of the ccenenchyma or with the axial portions of 

 the rectilinear and curvilinear fibre respectively, which, as 

 just stated, also gives evidence of similar hollowness, I am 

 not able to say; but hereafter I shall show that the rods were 

 probably closed where they projected beyond the free or 

 natural surface. 



The composition of all the fibre in the ccenenchyma, too, 

 always appears to be granular, and so open in some parts 

 that it appears almost cribriform. For this I have endea- 



