206 PEOF. H. A. NICUOLSOIf AND DE. J. MUEIE ON THE 



eminences. Some of them are apparently broken radial pillars, 

 and perhaps they are in general incipient structures of this nature. 

 Most of these tubercles are clearly imperforate ; but it is not 

 uncommon in silicified examples, as before remarked, to find a 

 variable number of the tubercles perforated by a central tube. If 

 these tubes are natural, then they must serve to place the succes- 

 sive interlamioar spaces in communication ; but, as above stated, 

 we have only detected them in silicified examples, and we are in- 

 clined to think that they are due to imperfect silicification. In 

 the tuberculate Stromatoporoids just alluded to (such as JS. granu- 

 lata, Nich.), whether we take calcareous or silicified examples, and 

 however exquisite may be the state of preservation, we have 

 hitherto failed to satisfy ourselves of the existence of any open- 

 ings piercing the laminae between the tubercles, and placing suc- 

 cessive interlaminar spaces in communication, though it seems 

 most likely that such really exist. 



On the other hand, there are forms (such as Stromatopora no- 

 dulata, Nich., and Byrincjostroma densa, Nich., and various forms 

 from the Devonian limestones of Devonshire and of the Eifel) in 

 which the surfaces of the horizontal laminseare smooth and exhibit 

 no tubercles, while they are rendered minutely porous by the 

 presence of innumerable apertures. These openings may be 

 extremely minute, or they may be of comparatively large size and 

 of a vermiculate character, in which latter case the horizontal 

 laminae assume a kind of trabecular or loosely reticulate structure. 

 There cannot, therefore, be in these cases any doubt that the 

 interlaminar spaces communicate with one another directly by 

 canals passing vertically through the horizontal laminae, and en- 

 tirely independent of the radial pillars. That some such com- 

 munication, to a greater or less extent, exists in all the Stromato- 

 poroids appears to be exceedingly probable, though we have 

 failed to satisfy ourselves of its existence in some forms. One 

 of the difficulties connected with this is that of demonstrating 

 any perpendicular tubules passing through the horizontal laminae 

 m transparent vertical slices of the fossil. Sections of this 

 nature ought to show these tubuli if they are present at all. In 

 very many cases, especially in unaltered calcareous specimens, 

 the substance of the horizontal laminae in very thin sections be- 

 comes so ill-defined that, even by the use of the polariscope, it is 

 impossible to make out whether tubules are present or not. Even 

 in cases where we know that the interlaminar spaces communicate 



