Madreporarian Genus Turbiiiaria. 519 



o( Astrcropora ; it differs chiefly in its greater abundance and 

 in its distribution, both of these depending upon tlie method 

 of budding. 



Rapidly growing cojnenchyma is always finely reticulate 

 or spongy; where streaming takes place it is often furrowed 

 in the direction of the streaming. The relative l)readths of the 

 furrows and of the separating ridges are indicative of the 

 density of the cocnenchyma. When the furrows are narrow 

 and the ridges thick and solid the coenenchyma is very dense. 

 The lower portions of a corallum are, as a rule, the densest, 

 and this can be traced by the ever-increasing thickness of the 

 ridges and corresponding diminution in width of the furrows. 



It is apparent, then, that before any systematic characters 

 can be based upon the coenenchyma its physiology requires to 

 be understood ; for instance, the spongy texture of that of 

 the margin of the cup is often given as a character, whereas 

 it is an invariable rule that in all such rapidly growing 

 portions the cojnenchyraa is spongy. 



Tiiere are, however, peculiarities which are to be noted, 

 viz. the characters of the trabecula3 building up the coenen- 

 chyma. They may be filamentous or lamellate, giving in the 

 former case a spongy, in the latter a flaky appearance to the 

 ccenenchyma, or they may be close and granulated, making 

 the surface look like sandpaper. In others, again, the ridges 

 are continuous, i. e. only broken by pores between the neigh- 

 bouring canals at long intervals ; in others the ridges are 

 highly echinulate, even broken up into rows of points repre- 

 senting so many open communications between the furrows, 

 which, when covered over, will become canals. 



Further, a certain value may be put upon the fineness or 

 coarseness of the texture. In some it requires a glass to see 

 it at all, in others it is visible to the naked eye. 



In connexion with this subject of the general aspect of the 

 coenenchyma, it is worth noting that this seems to vary with 

 geographical position. There are groups of specimens from 

 various parts of the world evidently in each case collected at 

 the same time and from nearly the same spot. In each case 

 all the specimens of these groups look at first sight strangely 

 alike. Tiiis is notably the case with a group from Formosa, 

 with another from Tongatabu collected by J. J. Lister, and 

 with another from Shark's Bay collected by Sa>ville Kent. 

 So strong is the likeness between the specimens in each case, 

 that without some definiti; principles of classification one could 

 hardly avoid lumping them all together, as, indeed, I found 

 had been done with the Formosa specimens. It was only 

 wiien, little by little, the different methods of growth and 



