Affinities of Astrosclera willeyana, Lis'er. 381 



The walls of the tubes were beset with siliceous spined 

 acanthostyles set at a very acute angle and evidently 

 pointing in the direction of exhalant currents. The spicules 

 beneath the ectosorne were vertically dressed. 



The gaps in the tubular network, i. e. the meshes left after 

 decalcification, were filled, in the complete state, with spherules 

 of aragonite, separate in the upper part of the sponge, but 

 fused into solid walls and masses in the lower. 



All four specimens have the siliceous acanthostyles. One 

 specimen has in addition cent rotria3iies, spined microxeas, and 

 spii asters belonging to a species of Tiipiolemus, a Tetracti- 

 nellid Theneid sponge, which seems to have a boring habit. 

 The flagellated chambers are fairly well preserved. In 

 some double-stained sections not thoroughly decalcified the 

 aragonite spherules took the stain, the deeper stained central 

 part of each spherule looking like a nucleus. Numerous 

 embryos are present, especially below the ectosorne. 



In 1900 J. J. Lister published a memoir on Astrosclera in 

 Willey's ' Zoological Results,' part iv. Four of the specimens 

 he described came from 35 fathoms off Lifu; a fifth from 

 100 fathoms off Funafuti was sent to him from the British 

 Museum. 



He considered these organisms to be very aberrant calca- 

 reous sponges which appeared to show affinities with some of 

 the Pharetronidse. Recently I wrote to Mr. Lister asking 

 for the loan of his preparations, all of which he very 

 kindly sent to me. 



In one of the longitudinal sections of a complete specimen 

 showing hard and soft tissues I found on the outer edge of 

 the section a number of vertically dressed acanthostyles, but 

 the tissues of the canals were devoid of these spicules. 



On decalcifying a fragment of the old and dead Funafuti 

 specimen I found remains of much macerated acanthostyles, 

 often reduced to a thin shell, but still recognizable as acantho- 

 styles. 



Mr. Lister had had imperfect material to work with, and 

 could hardly have come to any other conclusion than the one 

 he adopted. 



My own views concerning this sponge are as follows : — 

 Astrosclera willeyana is a siliceous Ectyonine sponge related 

 to Hymerhaphia. Originally it was a small thin incrusting 

 .'ponge with a skeleton constructed of separate vertical 

 columns of acanthostyles with one spicule to each column. 

 It acquired the habit common among its Ectyonine congeners 

 of selecting special kinds of particles from its surroundings, 

 which here include numerous aragonite-forming algaj corals, 



