﻿TUBERCULATE MONTIPOE^. 145 



immersed or else with thin, irregular, warty margins, formed by minute, solid, granular, coenen- 

 chymatous swellings (tubercles) of different sizes. 



A section of one of the thicker branches shows a solid layer in contact with the worm- 

 tube or tubes, surrounded by a dense reticular layer and a trabecular layer (not developed in 

 the thinner branches). The trabeculse are usually very stout and solid, and project above the 

 ccenenchyma as pronounced cylindrical granular or bushy tubercles, so large that they might 

 be mistaken for papilla. Without being regularly arranged round the calicles, they are often 

 taller near the polyp cavities. Eound the edges of the explanate portion the tubercles give 

 place to radiating ridges, and on the branches, especially near their tips, the tubercles fuse into 

 plates and ridges bent at all angles to one another, sometimes triradiate, sometimes thin and 

 even, at others swollen along their upper edges. Scattered among these plates are simple 

 cylindrical tubercles of all sizes, but mostly mere granular knobs. 



The specimen of M. effusa from Tahiti figured by Dana was a flat explanate growth with 

 slightly turned-up edges and one plate growing over another, and from its surface bent 

 branches arose abruptly as encrustations of worm-tubes. The smaller drawing (4a) also 

 clearly shows the surface to have been covered with tubercles ; while in the larger drawing 

 the fusions of these into bent plates and ridges is faintly but, it seems to me, unmistakably 

 indicated. The thickness of the branches is given as 16 mm. 



There are in all ten specimens in the National Collection which I place under this 

 heading. Of these, four consist essentially of bent and fused processes encrusting worm-tubes, 

 while four are almost purely explanate, with but very poor and feebly developed processes 

 due to worm-tubes. In all ten cases the tubercles are pronounced, and fuse to form ridges 

 and bent angular plates thickly covering the whole or parts of the surface. These specimens, 

 however, differ from one another so strikingly in general appearance, that it was only after a 

 long time, when revising and re-revising this attempted classification, that I eventually brought 

 myself to group them provisionally together. It is indeed worth expressing a doubt whether, 

 in view of the differences in the calicles and in the characters of the tubercles, this is a 

 species at all, and not rather an array of tuberculate Montipores accidentally but somewhat 

 similarly distorted by worm-tubes. 



The commensal worm-tubes wliich seem necessary to give this coral its typical branched 

 character are about 1 mm. in diameter, and occur singly or in groups. In one specimen (c) I 

 have counted thirteen in the section of a broken branch. In specimen b, whereas six tubes 

 are seen in a section of two fused cylinders, more than twenty-five open at the tips of their 

 branches. It appears, therefore, as if the worm-tubes must branch within the coraUum. Here 

 and there a tube in the centre of a branch does not open at the tip, the worm having either 

 died or been killed by the more rapid growth of the ccenenchyma encrusting it. 



a. The largest specimen from the Persian Gulf is a confused tangle of thin cylindrical 

 branches, each with a different curve (PL XXVII). Owing to the thinness of the ccenenchyma 

 encrusting the worm-tube or tubes, the trabeculse are only here and there slightly developed, 

 but the ridges and tubercles which, if the branches thickened, would become (when immersed 

 by the thickening ccenenchyma) the trabeculffi, show all the typical shapes. A great number 

 of sponges shelter among the interstices of the interlacing branches. The original explanate 

 portion has decayed away, but here and there fresh attempts to form explanate platforms can 

 be seen. Dark fawn colour. 



