444 



layer was still in a continuous sheet, and not folded to form chambers, 

 the thinness of the body-wall is just what might have been expected. 



There is thus concrete evidence from two distinct sources in support 

 of the theory of Hexactinellid phylogeny here advanced. It is of course 

 possible to discount the facts in both cases, and to explain them in other 

 Avays. With regard to the larval spicules, Ijima [3, III. p. 43] is of 

 opinion that "the stauractinic form of the spicules — in ontogeny is due 

 to the suppression of one of the three primitively present axes, in adapt- 

 ation to a certain secondary condition of the larva — assumably to 

 circumstances of the space in which the spicules develop themselves". 

 On fossil sponges there is no one whose opinion bears greater Aveight 

 than Dr. G. J. Hind e , and in regard to the question under consider- 

 ation he writes to me as follows: — "I do not think Schrammen's 

 sub-order Stauractinophora is very firmly based, for it is very probable 

 that in sponges of the families he includes in this group there are six- 

 rayed spicules as well as the cruciform ones. This has been pointed 

 out in descriptions of these sponges, in some instances small knobs or 

 prominences can be seen giving indications that the spicules were at 

 least five-rayed. But as most of these sponges are merely flattened im- 

 pressions in shale it is impracticable to ascertain whether the cruciform 

 spicules did or did not, in some instances at least, possess another axis 

 at right angles to the plane of compression. Some of the sponges de- 

 scribed by Dawson from the Ordovician of Metis, Quebec, (Trans. Roy. 

 Soc. Canada Vol. VII. 1889) give strong indications of having 5-rayed 

 spicules — whether a sixth ray was originally present is less evident. 

 Also in Hall & Clarke's Dictyospongidae it is stated that 'regular 

 hexactins' are in the parenchyma, and the figures on p. 30 show 'primary 

 hexactins or pentactins in place' apparently forming part of the »Stütz- 

 skelett« of the sponge. 



"At the same time there seems good evidence that cruciform 

 spicules did take a very prominent part in the skeleton of 

 these palaeozoic sponges [the italics are mine], though not more so 

 than the elongate rod-shaped spicules which Clarke says 'mostly com- 

 pose the vertical and horizontal bundles of the Dictyospongidae'", 



The statement just quoted concerning the stauractines is of the 

 greatest importance, for, as I may point out, it is by no means necessary, 

 on the theory here put forward ^-egarding the origin of the spicules in 

 these sponges, that all the palaeozoic forms should possess a skeleton 

 composed exclusively of stauractines; on the contrary we should rather 

 expect to find other forms of spicules occurring with them. Similarly, 

 there was probably a stage in the evolution of the Calcarea in which the 

 skeleton consisted entirely of monaxons, but such a condition is very rare 



