248 



Prof. P. M. Duncan on Thallophytes 



ments are onl j in the organic substance or are only surrounded by very thin 

 films of carbonate of lime ; for when these large tubes penetrate the solid 

 parts of the septa their diameter appears to be slightly less, their cell-wall 

 ceases to be distinguishable, and their contents are usually undivided. 

 But occasionally it appears that some burrow very successfully, and a 

 long and wide tube results, which, in the majority of instances, is full of 

 granular cytioplasm and rarely sufiiciently empty to exhibit a central 

 bright band of transmitted light. 



The Minute Filaments f rom tlie large hinds. — Although the broad fila- 

 ments found within the septa (especially of Caryoi:)hijllia claims, var. 

 horealis, of Flahellum laciniatum, Edw. & H., and of Balanojohyllia verru- 

 caria) do not often bifurcate or ramify, yet they now and then gi^'e off 

 long ramifying but not often anastomosing filaments of extreme tenuity. 

 They are found in crowds in some spots of the septa, and their diameter 

 is about YoAfoo ijich, their general distribution and method of branching 

 being a very exact counterpart of some of the larger growths already 

 noticed as occurring so frequently in the other hard structures of corals. 



Usually these very fine canals retain the same diameter throughout 

 their length. Li some corals they do not run a long course, but bend 

 soon or even ramify abruptly, and in both cases pass into irregularly 

 shaped cavities, emerging from them at the opposite end or absolutely 

 terminating in them. These cavities are not unlike ill-developed bone- 

 lacunse in appearance, and are frequently found in series on the periphery 

 of masses of long, radiating, spicular structures. Usually of very small 

 size, and more or less irregular in shape or elongate, they sometimes are 

 the cut ends or tubes of irregularly shaped parasitic borings which have 

 been formed by another and larger filament. 



The resemblance of these fine canals to those which pass off so fre- 

 quently from the larger parasitic tubules in the thick wall or columella 

 of the same or other species of coral is very exact, and probably they 

 have the same cause of origin. 



Large confervoid-looh'ng Filaments in the Body of Corals. — In trans- 

 verse sections of some corals, especially in the genera Lojfhohelia and 

 Solenosmilia, some long rows of the very large confervoid-looking 

 filaments are often seen cut across more or less obliquely. These rows 

 are formed of numbers of large tubes placed side by side and in one or 

 more layers, and they occur in those situations where involutions of the 

 dermal structures took place during the growth of the coral. These 

 tubes give out fine ramifying tubes such as those just mentioned, and 

 which usually terminate in irregular-shaped cavities. 



But usually the contents of the large tubes are crowded and not sepa- 

 rated off, and it is rare to see spaces in them through which light can 

 penetrate. They branch, and usually produce ramifications which are of 

 the same diameter as the parent filament ; and where they are cut across 

 they present an irregular granular appearance, which is partly due to 



