387 



cylindrical perforations in the rock whicli they at one time filled, while 

 the existence of the stolons is only indicated by grooves such as those 

 represented in figs. 359, 362. These facts seeem to prove clearly that 

 the space between the tubes was not filled with shell substance, but either 

 empty, or entirely, or partly full of soft matter, which was immediately 

 dissipated after the death of the animal, and its place occupied by the 

 soft mud in or on which the creature lived. Were it otherwise, we would 

 now find the space in question a compact mass of calcareous spar or amor- 

 phous silex, while the tubes (or cells as they would be in that case) might 

 be filled with limestone. In the magnesian specimens the ectorhin seldom, 

 if ever, remains ; and in species with flat plates the form (of the plates) 

 can rarely be made out, the only markings on the surface being the 

 grooves of the stolons. But where the plates were deeply concave the 

 position of the sutures is indicated by more or less strongly elevated ridges 

 enclosing rhomboidal depressed spaces with a tube cavity in the centre. 

 Fig. 362 represents a fragment of R. Ccmadensis in that state of pre- 

 servation. The rhomboids in this case are not the plates themselves, but 

 only their impressions. In describing such specimens, the tubes are 

 sometimes spoken of as having rhomboidal openings, but this is an error ; 

 the tubes when perfect, as can be proved by hundreds of specimens, are 

 not open at all, but completely closed, at one end by the ectorhin and at 

 the other by the endorhin. They all, however, communicate with each 

 other through the stolons and endorhinal canals. 



Were the tubes of Receptaculltes to be closely crowded together so that 

 their walls would everywhere be in contact, and no space between them, 

 then the structure would be similar to that of OrbitoKtes, but with the 

 system of connecting stolons arranged on a different plan. The genus 

 would then also closely resemble Dactyhpora ; but I do not yet see that 

 the evidence is sufficient to prove clearly that the tubes are strictly the 

 homologues of the cells of any group of the Foraminifera. They appear 

 to me to be more nearly related to the connecting spicula of the Spongidge. 

 Each tube with its cylindrical shaft, and plate at each extremity, resembles 

 not remotely a birotulate spiculum. Or it might perhaps with more pro- 

 bability be described as consisting of two spicula united at their points. 

 Thus the ectorhinal plate with the four stolons may be a peculiar form of 

 the foliato-peltate spicula, of which many different kinds are figured by 

 Bowerbank. The cylindrical shaft may be a spiculum approaching the 

 acuate or acerate varieties with its point inserted into the nucleus of the 

 foliato-peltate spiculum. Most sponge spicula are hollow ; and we know 

 how often it happens in the structure of the animal kingdom that organs 

 may at one time subserve one function, and elsewhere a very difterent 

 function. The cylindrical cavity, which in the spicula of the ordinary 



