ACANTHASCIN^. 131 



armed with prongs, are often, though not always, paratropal, i.e., 

 they are, as it were, pushed to one side in their plane so that 

 they form with one another three acute angles and one obtuse 

 angle greater than 90° or even 180°. Similar hypodermal pentactins 

 have long been known in llossella antarctica. As in this species 

 they are generally found in groups of several together. In every 

 such group, the proximally directed shafts, accompanied by 

 comitalia, form a more or less compact column or tuft that dips 

 deep into the choanosome, while the four-rayed heads produce on 

 the sponge surface a star-like figure in which a number of streaks, 

 i.e., the paratangentials, radiate in all directions from a central 

 space (PI. XIII., fig. 12 ; Pi. XVIIL, fig. 16 ; PL XIX., fig. 

 23 ; PI. XXIL, fig. 10 ; etc.). It is easy to discover that the 

 separate pentactin-heads in a hypodermal group lie one above 

 another in close order, the upper one of any two being older and 

 more fully developed than that next below (see PI. XVIII., fig. 

 16 ; etc.). The lowest is therefore tlie youngest, which develoj)S 

 while clasping in one of its angles the column of shafts belonging 

 to older pentactins. It is this preexisting shaft-column that 

 disturbs the regularly cruciate development of the paratangentials, 

 forcing these to deviate sidew^ays from their normal directions ; 

 hence, their paratropal arrangement. In Rhabdocalyptus it is in 

 the older pentactin-heads only that the rays are armed with 

 prongs ; in the younger and therefore the more deeply situated 

 heads the rays are smooth. 



The hypodermal pentactins remain in the locus nascendi only 

 in certain species. More often they are destined to be protruded 

 outwards through the dermal laver and thus to form a second 

 kind of pleural i)rostalia, the other being the diactinic prostalia 

 before mentioned. The pentactinic prostalia stand out isolated or 



