DISTRIBUTION OF GUiOWTH-FORMS. 137 



P. West Indies x. 18 (PI. XIV. fig. 4). Tends to form mas.sive flabellate base, with 

 fringe of tall erect branches of irregular thickness, which fork again, so as to 

 form fresh flabellate bases. 



P. West Indies x. 18 (PI. XIV. fig. 5). The tips swell and divide into rounded knobs 

 very irregularly ; stems keep as nearly as possible erect. 



P. West Indies x. 19 (PI. XVI. fig. 1). A dense cluster of irregular stems with knob- 

 like tips ; stock expands into an inverted cone, upon a thin basal stem. 



P. West Indies x. W (PI. X. fig. 2). Thin, short, wavy, nearly erect stems, flattening 

 and swelling irregularly. 



APPENDIX TO TABLE III. 



On the Growth-Form of Porites x. 10. 



Among the Porites from unknown localities, there is one growth-form which demands 

 special notice, viz., that of Porites x. 10 (see PI. XII. fig. 2, and PL XVII. fig. 12). ^ It is quite 

 unique in the genus so far as our personal knowledge goes. It is a very pronounced coenen- 

 chymatous form showing some ccenenchymatous papillae, like a papillate Montipora. This, 

 I think, definitely establishes it as an Indo-Pacific form, for the ccenenchymatous Porites are 

 so far unknown in the Atlantic and West Indian area. It appears to have grown out from the 

 side of some object as a large irregularly ear-shaped dish, shallow with very thin delicate edges 

 and about 12 mm. tliick where it was attached. In order to bring this into line as a product 

 of a metameric series of colonial units, we have to conceive of the series as a pack of cards 

 lying horizontally on one another, with a certain number superposed directly upon one another, 

 thereby thickening the lateral attachment, while above this the successive growths are pushed 

 further and further outwards, so as to form the dish or cup described. This was not all, for 

 we can gathet two other important points : (1) the new growths start from the edges, that is the 

 highest point, and grow backwards, often perhaps arching over the old stock ; and (2) this kind 

 of growth shows us that the successive growths have some general average size, as we should 

 expect them to have if our supposition that they are metameric repetitions is correct. For the 

 new growth upon the edge tends sooner or later to break up into lobes, as if no single colonial 

 unit is large enough to cover the whole edge, after, in the expansion of the cup, it has exceeded 

 a certain size. 



