14 



For extemal appearance and general anatomy, see p. 2 — 6. 



S p i cu 1 a. 



1. Tylostyle. (PI. IV, fig. ia— 5; fig. 2a— J; fig. 6a— £; fig. 8a— e). On the whole the 

 tylostyles of P. melobesioides resemble those of P. carinata. We can likewise distinguish cortical 

 and parenchymal tylostyles, and also terminate the latter more obtusely (Cf. PI. IV, fig. 2). 

 For the size (in u) the following table may be sufficiënt. 



As in the case of P. carinata, previous authors did not distinguish cortical and 

 parenchymal tylostyles, and the measurements of the tylostyles in general are, therefore, of 

 little value for comparison with our observations. 



2. Sterrospira. (PI. IV, fig. 1 t— ).; fig. 5 a— £; fig. 70c, ?— ïj, t; PI. V, fig. 2—4; fig. 6; fig. 10; 

 fig. 12). We have seen in our general account of the sterrospirae how far they are 

 distinguished from those of P. carinata. The greater variety in size which we pointed out 

 p. 12 is best seen by comparing fig. 10 with fig. 12 on PI. V. 



3. Spherula. (PI. IV, fig. 3; fig. 4; fig. 6>j. PI. V, figs. 10, 12). Although variable in size as 

 fig. 3 demonstrates, the spherules remain always very minute; they are smooth, roughened or even 

 spined; in one case (1849 a) they often seemed to have the shape ofan asterisk (PI. IV, fig. 6vj). 

 They replace the microspires of P. carinata ; like them they are typically dermal, lining the 

 external sponge surface and the canals in the way described above for the microspires. 

 It is a very remarkable fact that they occur abundantly in the parenchyma; not separately 

 however, but in groups (PI. V, figs. 13 and 14), as observed already by previous authors. 

 We are inclined to believe that such groups represent early stages of development 



- perhaps they originate in a single cell • ■ and are afterwards transported towards the 

 periphery or the canals where they become isolated. 



4. Spheraster. (PI. IV, fig. 7 {3; fig. 8 3- and /.. PI. V, fig. 10). Exactly as in P. carinata the 

 spherasters in P. melobesioides are rare or absent. They are absent - - at least we iailed 

 to find them in 1849 b and 1033; on the other hand they are relatively frequent in 

 660 a and 660 b. 



