662 E. KIRKPATKICK. 



foundation on which it is growing. This close union is one 

 of the characters that has made tlio problem of Merlia 

 difficult to solve, aud has led to false inferences being drawn, 

 as I shall endeavour to explain later. 



In addition to the above-mentioned specimens there are two 

 much larger ones requiring special notice. One of these, a fine 

 example forming a large patch over 30 square centimetres in 

 area on a small block of volcanic i-ock, was hooked up by a 

 fisherman from ninety fathoms off Porto Santo Island (PI. 32, 

 fig. 4). The specimen was nearly di-y when brought to me, but 

 it still retained its red colour. Lastly, in the Seminario 

 Museum at Madeira there is a large fragment of a dead 

 Dendrophyllia covered over a considerable area with a thin 

 crust of Merlia. The coral was hooked up from ninety 

 fathoms off Cape Garajau, Madeira. PI. 32, figs. 5, 5a shows a 

 small piece of incrusted coral, which the authorities of the 

 Seminario Museum kindly allowed me to brealc olf. The 

 Merlia incrusts both the sides and free end of the coral 

 branch. Evidently the sponge flourishes better in ninety 

 fathoms than in sixty, and it may be interred that the habit 

 of Merlia is always incrusting. 



When living specimens are removed from the sea the soft, 

 smooth surface sinks down a little, thus allowing the porcelain- 

 like network, with its minute circular or polygonal meshes 

 and little nodal tubercles, to be seen imbedded in the red 

 flesh. Usually no openings of any kind are to be seen, but 

 in the case of two specimens killed suddenly by dropping 

 alive into Flemming's mixture, several very small circular or 

 oval holes, from 20 to 60 ju in diameter, were visible (PI. 32, 

 figs. 7, 8). These two specimens were the best preserved of 

 all, for they were in the expanded condition, and the fixing 

 fluid had penetrated well into the interior. 



The thin growing edge of the sponge extends well be3'ond 

 the edge of the calcareous skeleton in examples in which the 

 crust spreads freely over a surface, but often the growing or 

 creeping edge is dammed up, so to speak, and the sponge 

 and its calcareous skeleton increase in depth at such places. 



