78 



subject to gient variation in occurrence. This does not exclude the possibility 

 however that in a number of cases they may be good auxiliary characters. As 

 a rule the pores are absent in the division Maktcoslega. where they no doubt are 

 replaced by the membranous frontal area. In most species of the genus Electra 

 however, we find pore-like spots which in reality are only thin places in the 

 gymnocysl. On the other hand, we find pores in most other CheUosloinnla, but 

 they are constantly absent in the zotrcia of the genera Cellulariu, Hippothoa, Chorizo- 

 pora and Eiithyroicles. They appear generally only on the frontal wall, and on the 

 basal wall only in the families Eiithi]ridae and Petraliidae, the latter family of 

 which mainly embraces free forms with one layer, and the basal wall is very 

 often furnished with one or a few pore-chambers, more rarely with scattered pores. 

 The pores on the frontal wall may appear as marginal pores, or as scattered; 

 but this separation is not always sharp, because the marginal pores may appear 

 in several rows, and thus gradually extend over a larger or smaller part of the 

 surface, and cpiite apart from this transitional state the ])ores in several species 

 may appear sometimes as marginal sometimes as scattered pores. Tliis is for in- 

 stance the case in Porella concinnu, Escharina Hijudmanm and Haplopoma impressa. 

 We may cite the Steganoporellidae, Thalamoporellidae, Petraliidae, Hippnpodin'uUte 

 and Microporelld as examples of families and genera in which scattered pores 

 appear constantly, while the pores are more variable in their occurrence within 

 the genera Escharello, Escharoides and Smitlina, as well as in the family Adeoni- 

 dae. In the families Reteporidae, Celleporidae and HoloporeUidae we find as a rule 

 a verj' small number of pores, which are mostly situated on the marginal por- 

 tion of the zooecium, and in the family Onchoporidae we have external rosette- 

 plates appearing in small number in the distal half of the zoo'cium. 



The spines in regard to their systematic importance may be compared with 

 the pores, and their occurrence shows a similar lack of constancy. From their 

 shape and structure we can distinguish between jointed and nnjointed, single and 

 branched spines. The joints seem only to be a practical arrangement, to enable 

 longer spines to better resist pressure and blows, and as a rule consist in the 

 spine at the proximal part being I'lirnislied with a chitinized belt which gives it 

 a certain flexibility. In specially long spines such joints may be repeated up to 

 nine limes, and such articulated spines have thus a certain resemblance to the 

 antenna' of many insects. They occur rather seldom and are only found in a 

 number of Retepora species, in Escharello diapluina (Fl. XVII, fig. 1 a) as also in 

 the species of the genus Exochella. All short spines are on llie oilier liarxl nn- 

 jointed. More or less strongly branched spines are also very rare and can be 

 found in single, double or still larger numbers in various families and genera. 



