AFPrNITIES OP THE GENUS PROTOSPOXGIA. 365 



the secondary rays, and one by a primary ray. [The imaginary 

 square of the primary spicule may have been, aDd probably was, 

 equally actual ; but in the specimen before us there is nothing to 

 show this.] 



The rays of the tertiary spicules end by passing beneath the rays 

 of the primary or secondary spicules, just as the rays of the secon- 

 dary passed beneath those of the primary spicules. 



The quaternary spicules are arranged in like manner, their centres 

 lying in the centres of the squares formed by the other spicules, and 

 their rays being directed at right angles to the other rays, beneath 

 which they terminate. Sometimes, however, one or more rays of a 

 quaternary spicule overlap, while the remainder uuderlap, the rays 

 of the surrounding spicules ; but this is quite exceptional. The 

 spicules are thus symmetrically arranged, their rays all lying regu- 

 larly disposed in two directions at right angles to each other, and 

 so building up a network with square meshes. The following 

 figure (fig. 2) will serve as a diagram of this arrangement. 



Pig. 2. — A single large spicule of Protospongia, with the smaller 

 spicules filling its interspaces. 



One notices in Dr. Hicks's specimen a tendency to depart from 

 strict rectangularity in the meshes, through a general approach of 

 the spicules at one end of the specimen and a general divergence at 

 the other, as though the part of the skeleton exposed had once be- 

 longed to a more or less spherical sponge, and not to one having the 

 form of a mere flat film or plate. The symmetry of the arrange- 

 ment is also liable to be disturbed by many minor irregularities, a few 

 of the smaller spicules being more or less difl'erently oriented from the 

 remainder. In one case also the centre of a quaternary spicule ap- 

 pears to be seated almost exactly upon the ray of one of the larger 

 forms, from which one might infer that this particular spicule, at all 

 events, did not posses a fifth ray. 



Thickness of the Sponge-wall. — As regards this we have as 

 yet no certain knowledge ; but it appears to have consisted of 

 more than one layer of spicules. In the specimen two layers 

 of sponge-structure are shown at different levels, separated by a 

 little "cliff." But not much is to be concluded from this ; for the ' clifi"' 

 may be merely due to a ' slip ' or minute fault which has shifted an 

 originally single layer of sponge-structure to diff'erent levels ; or even 

 if this be not the case (and the presence of a layer of iron-oxide 

 continuing the direction of the lower layer beneath the upper seems 

 to show that it is not), there still remain three possible interpre- 

 tations of the double layer : the sponge, for instance, may have 

 been thin-walled and sacciform, and its collapse may have brought 



