462 ASTROSCLERA WILLEYANA, THE TYPE OF A NEW FAMILY OF SPONGES. 



It is further to be observed that the trend of growth of the later formed bosses 

 is in a direction deflected at an angle of about 60° from the axis of the original 

 growth. We may conjecture that this result has been produced by the shifting of 

 the position of the object — block of coral or what-not — to which the specimen was 

 attached. 



The upper surface, whether of the original disc or of the later formed bosses, is 

 perforated by closely set pores (Fig. A, 3 and 4). These are sometimes isolated, but 

 often they open into curving and branching grooves, recalling those of the coral 

 Maeandrina. At seven places on the surface of the later growth, and at one on the 

 original growth, the lines of pores are seen to be disposed in a radiating manner 

 about so many centres. The lines about a centre vary in number from five to ten, 

 and at their outer ends they become lost among the smaller pores round about. 

 The pores along these lines are large (Fig. A, 4), being at least twice the diameter 

 of those distributed elsewhere over the surface, and a group of such pores is situated at 

 each of the centres from which the lines radiate. 



There is then a differentiation of the pores, and hence of the canals, whose openings 

 they are, into two categories, viz. large ones disposed in radiating systems, and small 

 ones distributed between the systems. As will be seen later, the distribution of the 

 pores at the surface is the expression of an arrangement of the canals found throughout 

 the interior. 



The soft parts proper to the organism were not present in the portions of this 

 specimen that I have decalcified, and a boring sponge, with characteristic pin-shaped 

 siliceous spicules and bundles of raphides, appears in the sections. Moreover, the surface 

 even of the most recently formed part is dotted over with the pink growths of the 

 Forarainiferan Polytrema, a brown alga has established itself at one point (these have 

 been omitted in the figures), and the base has been much excavated by boring organisms. 

 It appears, therefore, that this specimen of Astrosclera had died some time before 

 it was collected. 



Methods. When collected by Dr Willey the Lifu specimens were put into strong 

 alcohol (over 70 °/„). In examining them two have been decalcified by means of dilute 

 acetic acid, added drop by drop to a dish of 70 °/^ alcohol containing them, until bubbles 

 of gas appeared. When decalcification was completed, the soft organic body was divided 

 longitudinally, and the parts embedded in paraffin and cut into sections — one half 

 into longitudinal sections, the other into transverse. Another specimen was dealt with 

 as follows. It was first divided longitudinally by a fret-saw, to allow the reagents to 

 penetrate the interior more readily. Some idea of the degree of hardness of the 

 skeleton may be gained from the fact that the lines of the sawing were sharply 

 marked on the cut surfaces. On the other hand, it was not so hard as to perceptibly 

 blunt the saw, a result which was soon effected by the skeleton of a specimen of 

 Polytrema, on which I made a preliminary experiment. The upper part of one half 

 of the specimen was sawn off", decalcified, and cut into sections, for comparison with 

 the sections of the other specimens. The remaining parts were then stained in borax 

 carmine for two days, passed through 90 °/„ and absolute alcohol into chloroform, and 



