446 



whatever. If, on the contrary, the folding of the collar-cell layer was 

 brought about in order to increase its superficial extent, as seems very 

 likely, it might rather be supi)Osedthat the diverticula would tend to take 

 on an arrangement hke the cells of a honeycomb (fig. 2 ò), since such an 

 arrangement gives, as is well known, a much greater economy of space, 

 and would allow of the diverticula being much more closely packed. If, 

 however, the diverticula had the honeycomb arrangement, and the 

 spicules were adapted to this pattern, the result would not be a staurac- 

 tine or other form of spicule with rays meeting at right angles, but a tri- 

 radiate spicule, as in Calcarea, with the rays meeting at angles of 120°. 

 It is my belief, therefore, that the disposition of the folds or chambers 

 was not the cause of the rectangular junctions of the tangentially placed 

 spicule-rays, but that, on the contrary, the possession of regular staur- 

 actines was an antecedent condition, and that the folds or diverticula 

 of the gastral layer owed their peculiar arrangement to the fact that 



dosò- Ä» 





Fig. 2. Diagrams to show two possible arrangements of the chambers or diverticula, 



as seen in tangential sections of the body-wall, and the forms of spicule that would 



result from adaptation to the arrangement in each case. 



they gi'ew out into the rectangular meshes of a skeleton composed of 

 stauractines ; just such a skeleton, in fact, as is sliOAvn in Sollas's figure 

 of Protospongia. It is not improbable that the folds of the gastral epithe- 

 lium in staged arose pari passu with the development of radially directed 

 rays upon the stauractines, just as in Ascandm falcata the folding of 

 the gastral layer is correlated apparently with the great development of 

 the gastral rays of the quadriradiates. We are then forced to seek some 

 other cause for the symmetry of the stauractine, and there are only two 

 possible explanations which suggest themselves to me. One is that the 

 form of the stauractine was due to a rectangular arrangement of the 

 meshes of the trabecular framework in which it arose. There is however, 

 not a particle of evidence to show that the trabecular framework has, 

 or ever had, such a regular arrangement of its constituent trabeculae as 



