row of plates on each side of the middle one has one half within the same zooecium, 

 and the other half within the adjoining zooecium. These apparently half parts do 

 not belong to each other however, as each of them passes over into and joins at 

 a right angle to one of the small plates of which each latei-al wall consists. That 

 two adjoining rows of half plates seen from the surface of the colony can look 

 like a row of whole plates, is due partly to the fact that they fit exactly ot each 

 other, partly that from the surface we cannot see the halves adjoining one another 

 in the vertical walls. Further, as the zooecia in one layer of the colony alternate 

 with those in the second layer, a median row of plates in one layer will corre- 

 spond to a double row of adjoining half plates in the other, and the two ad- 

 joining half plates are in size almost exactly like the opposite plate. The ter- 

 minal wall is divided into two lateral halves by a suture, running medially through 

 the single rosette-plate. At the borders between the single longitudinal rows we 

 also find small uncalcified interspaces, while the plates in the single rows are 

 separated by narrow sutures. I have in two previous papers^ designated this 

 manner of calcification as circular, because the calk particles in the individual 

 small plates are circularly arranged round a small condensed shining spot, which 

 we might call the ^centre of calcification*, and which in the angularly bent plates 

 is placed in the angle between the two pieces of each plate. The circular arrange- 

 ment is most distinct close to the centre and vanishes gradually further out. 

 Still, more or fewer zooecia show a less regular arrangement of the small plates 

 in the marginal part of the colony, and the same can be seen in scattered zooecia 

 in other parts of the colony. The basal surface may then either be broken up 

 into an irregular mosaic of larger or smaller plates of different shape, or the 

 median row of plates may be missing or represented only by very few plates. 

 We can even here and there find a cell-mosaic. I have found a plate-mosaic 

 like this on the lateral walls in Flustra foliacea (PI. XIX, fig. 9 b), on the basal 

 wall in Porella saccata, Por. compressa, Smittina trispinosa, Sin. palmata (PI. XIX, 

 fig. 5 b), Sm. linearis, Discopora pauonella, Flustra serrulata, Fl. pisciformis as well as 

 on the front wall of Inversiiila inversa and Anarthropora monodon. In contrast to 

 what occurs in Fl. securifrons none of these species show a regular arrangement 

 of the small plates, and in a number of them the latter appear in a very irreg- 

 ular and variable way, as they may appear together with other forms of calci- 

 fication within the same colony, even on the same wall. Time does not allow 

 me to enter into details, but I will ju.st mention Smittina trispinosa, Flustra serru- 

 lata and Fl. pisciformis as examples of such species. While the small plates in 



' 54, p. 246; 55, p. 3. 



