on Plate IV we see that in the former case the centre of the rod is almost destitute of spines, 

 whilst these reach already a considerable length at the two extremities; whereas in the second 

 case the rod is almost equally spined over its whole length. This difference in spinulation may 

 be observed not only in the very first stages, but also later (PI. V, fig. i {3 — $). We have, 

 for some time, believed, that this was something of a specific character, as the terminal 

 spinulation is specially and very distinctly seen in P. carinata. But on the other hand it must 

 be stated that in P. carinata equal spinulation likewise occurs (PI. IV, fig. 9}. — 7). 



We have seen that, according to Keller, the early stages of the sterrospirae are 

 " spirasters," or "gekrümmte, dornige Stabe;" and Hanitsch says that the axis is "nearly 

 straight or slightly twisted." Evidently the twisted character is best seen in very early stages; 

 the more the spines grow and fuse by soldering silica, the more difficult is it to observe the 

 twist. And yet it can be observed in many cases even of almost fullgrown spicula. This can, 

 however, better be seen in the preparations by allowing the spicula to roll slightly, than in 

 drawings. [Cf. PI. IV, fig. 9-; fig. 1 5 (- — ^- P1 - V , % ï ^ ^ Ü S- 2 \ % 3*; n g- 5]- In very 

 young stages it is easily seen, though the torsion is never great. Everybody will agree that 

 such spicula as illustrated on PI. IV, fig. 52, f;; fig. 10:, /j; PI. V, fig. 32; fig. 4, are, 

 according to the old nomenclature "spirasters," or, according to our terminology spinispirae. 

 There can hardly be any doubt that these are young sterrospirae, for in P. mclobesioidcs no adult 

 spinispirae occur and in P. carinata they are of another character than the adult "spinispirae." 

 Moreover, we find a great many gradual transitions from young sterrospirae, having the shape 

 of spinispirae, to the adult ones. Just as in adult spinispirae the twisted character is seen 

 sometimes more, sometimes less distinctly, simulating simple, curved or almost straight, spined 

 rods, so we find exactly the same in the spinispirae, which are but early stages of sterrospirae. 

 We consider the sterrospirae of Placospongia as modifications of spinispirae, as spicula belonging 

 to the group which one of us (Vosmaer 1. c. p. 105) has called 2-spiraxons, because in favourable 

 cases and very accurate focussing this is certainly the case. 



The remarkable appearance of the surface of adult sterrospirae is already observed by 

 Oscar Schmidt (1870 p. 72) and his observations are in the main confirmed by everybody. 

 That there is variation with regard to the dimensions of the free portion of the spines, is 

 certain. On PI. V, fig. 1 r, and S we have given outlines of two sterrospirae from one and the 

 same specimen (311 b). On the left hand side of the figures is drawn the surface view ; on the 

 right hand side in fig. 1 j is drawn the view a little lower than the tops of the spines, i. e. 

 about at their base. This is, however, only seen in some cases. The explanation of these 

 different views seems to be this, that between higher, free portions of spines there are lower 

 ones. Focussing at the top the lower ones are not seen, except sometimes the shadow of the 

 sides, creating a network-design of smaller meshes within one larger mesh. Focussing at the 

 base, they are all seen as closely packed circles. The side view in such cases is then as in 

 fig. 1 3 on the right hand side. 



Somewhat different in shape are the sterrospirae of Physcaphora decorticans Hanitsch, 

 called by the author "selenasters," and those of Placospongia gracjfci Ldfd., called by the author 

 "pseudoster rasters." As far as we may judge from the statements of these two spongiologists, 



