124 DR. J. S. BOWERBANK ON THE SPONGIADjE. [Feb. 6, 



scribed in the account I have given of the construction of the 

 sandy fibres of Dysidea fragilis'vn my paper "On the Anatomy and 

 Physiology of the Spongiadse " in the * Philosophical Transactions 

 of "the Royal Society' for 1862 (pi. xxviii. figs. 3, 4, 5, p. 7o7)- 

 The mode of aggregation in the present case is well demonstrated in 

 some of the young primary fibres of the sponge. A finely pointed 

 purely keratose fibre is projected forwards, the termination of which 

 is adhesive, and to which any small body touching it becomes ce- 

 mented and is then speedily covered by a thin coat of keratode, the 

 surface of which is not adhesive. The adhesive point continues its 

 forward course continually, thus adding material to the arenaceous 

 axis ; while the successive developments of concentric keratose 

 layers, destitute of adhesive power, surround and maintain the 

 arenaceous axis in its proper position. The apparent discretion in 

 sizing the selection of granular matter may thus be naturally ac- 

 counted for ; and it may be readily imagined that the slender adhe- 

 sive advancing point of the fibre would not support and retain any 

 larger material than that which we find it to have appropriated. 

 The secondary fibres pullulating from the sides of the mature non- 

 viscid primary ones are not adhesive, and are therefore free from 

 arenaceous axes. 



The dermis is very thick ; and its strength is further increased by 

 a very abundant imbedment in its substance of particles of sand 

 and other small extraneous matters. This appears to be accom- 

 plished with great regularity, especially on its external surface, where 

 all the prominent ridges of the great reticulations are as closely and 

 regularly set with granules as if they had been the work of a lapi- 

 dary. The internal surface of the coriaceous dermis exhibits the 

 same structural peculiarities as the external one. 



The porous areas are not simple, as we find such organs in 

 other cases of their occurrence, where we have a thin pellucid mem- 

 brane with a group of three or four pores. In the present case the 

 porous area is large, thin, and transparent, and is more or less reti- 

 culated by fine thread-like lines of extraneous matter; and within each 

 of the little areas thus formed there is a single pore. It is a remark- 

 able circumstance that the boundaries of these minor areas are fre- 

 quently determined by triradiate spicula of calcareous sponges im- 

 bedded in the surface of the membrane, the radii of the spicula 

 forming more or less of the boundaries of three such areas. The great 

 porous areas are frequently oval and of considerable size, containing 

 as many as seven or eight of the small single pore-areas. 



The sarcode in this sponge is remarkable for its substance and 

 opacity ; the nearest familiar representative of it is the boiled albu- 

 men of an egg. When a thin portion of it is mounted in Canada 

 balsam and thus rendered transparent, a considerable number of com- 

 paratively large lentiform gemmules are to be seen imbedded in the 

 interstitial membranes. They contain a more or less well-defined 

 central nucleus with innumerable minute radial lines which reach to 

 the extreme margin of the gemmule. They are certainly not spi- 

 cula, but have every appearance of being minute tubuli. 



