42 PROCEEDINGS OP THE ACADEMY OF [1886. 



to explain by an incomplete silicification of the central portion 

 of these pillars and by subsequent lixiviation of the unchanged 

 calcareous centre on exposure to atmospheric influences. 



I have, likewise, frequently observed a central perforation of 

 cross-fractured pillars in calcareous, as well as in silicified speci- 

 mens, but while I perfectly agree with their representation of the 

 pillars as generally being solid, not tubular, I do not think that 

 the central perforation of the pillars in the mentioned cases is 

 the result of weathering of the imperfectly silicified skeleton 

 substance. I believe it is caused b}^ funnel-shaped inflexions of 

 the laminae at the base of these pillars, which excavation shows 

 itself as a central opening, if the apical part of the pillars which 

 is solid is broken off. 



The laminae of some forms of Stromatopora, such as tuberculafa, 

 and granulata of Nicholson, are described as continuous leaves, 

 without any visible perforations placing the superimposed inter- 

 laminar spaces in communication. It may be suggested that sucli 

 openings of communication may exist. I had frequently occasion 

 to observe this to be the case, although, in many instances, the 

 laminae of these species of Stromatopora appeared perfectly free 

 of larger perforations than the minute pores of the tissue itself. 



In other forms, such as Strom, nodulata, and densa, the exist- 

 ence of such larger openings for communication between the 

 superimposed interlaminar spaces has been ]"ecognized by the 

 writers. 



According to m}" own observations, a part of these pores of 

 communication, and at the same time the largest ones, can be deter- 

 mined to be portions of the ramified horizontal channel expansions 

 radially converging towards certain centres, and uniting there in 

 one larger central canal, ascending vertically. These horizontal 

 canals in their extension slowly ascend from one interlaminar 

 space into another. They have, as Messrs. Nicholson and Murie 

 correctly observed, no proper walls, and cannot be considered as 

 canals horizontally intersecting the substance of the laminae, but 

 with more propriety are described as furrows on the surface of 

 the laminae, which, finally, by the formation of a new lamina above 

 the older one, become roofed over and are transformed into closed 

 tubular spaces. This opinion I fully indorse. 



All Stromatoporas do not exhibit such radical channel expan- 

 sions well developed, but faint indications of radially arranged 



