28 PLATE XI. 



specimen of this species for examination, labelled 

 "56 miles east of Balta Shetland, 1867." This 

 specimen has thirteen fistula on an irregular mass of 

 old shells, &c., about one inch in diameter. Some of 

 these organs are long and slender, like that represented 

 in fig. 13, pi. x; but the greater number of them were 

 comparatively very much shorter and stouter in their 

 proportions. This specimen has the largest number 

 of cloacal fistula that I have yet seen on one basal 

 mass, and this fact, in combination with the shorter 

 and stouter proportions of those organs, causes it 

 strongly to resemble a specimen of P. mammellaris ; 

 and it is only by a microscopical examination that it 

 can be separated with certainty from that species. 



POLYMASTIA RADIOSA, Bowerbank. 



Vol. ii, p. 68, ' Mon. Brit. Spongiadse.' 



Figs. 14, 15. -The two type-specimens of P. radiosa. 

 Natural size. 



Fig. 16. Represents a portion of the surface of 

 fig. 15, exhibiting the large radiating groups of external 

 defensive spicula. X 80 linear. 



Figs. 17, 18. Two of the spicula from the radiating 

 groups of external, defensive spicula. X 123 linear. 



On January 15th, 1870, I received two more speci- 

 mens of this pretty little species, both attached side by 

 side, but separate from each other 011 the inner surface 

 of a triangular fragment of a bivalve shell scarcely 

 exceeding a superficial square half inch. These two 

 specimens were very little different from those figured ; 

 one of them was nearly of the same size and propor- 

 tions of fig. 14, pi. x. The other was of about the 

 same length, but broader in its proportions. Both 

 were attached to the shell for the whole of their length, 

 confirming in this respect the singular habit of the first 

 two specimens. Neither of these specimens exhibited 

 the beautiful radiose structure of the surface repre- 



