66 PROFESSOR OSKAR CARLGREN ON THE GENUS PORPONIA 



peculiar symmetry as that in Endocoelactidae may also be found in the free Actiniaria, 

 it seems reasonable to conclude that among the variously attached Madreporaria, with 

 their varying adaptation to the under-layer, still more complicated and varying 

 arrangement of the mesenteries may be found. In my opinion, the arrangement of the 

 mesenteries in Endocoelactidae opens up the possibility of a more intimate connection 

 between Rugosa and Madreporaria, and more readily leads to an explanation of the 

 conditions of symmetry in Rugosa, like the one proposed by me in Bronn's Klassen 

 und Ordnungen, p. 1 50. Whilst the development of the mesenteries in the Endocoe- 

 lactidse gives a greater possibility of interpreting Rugosa, it makes the question still 

 more complicated, as in Rugosa there might be a development of mesenteries in the 

 endocoels. Though we shall naturally never be able to reach finality with regard to 

 the position of Rugosa as compared with the typical Madreporaria, but have to be satis- 

 fied with a hypothetical explanation, so long as we do not know how the mesenterial 

 musculature is arranged, I shall nevertheless give a picture of the way in which we 

 might imagine the origin of the Rugosa type, if the mesenteries after the 6-pair stage 

 have developed in the endocoels. I presuppose that the hypothetical, separating walls, 

 sarcosepta, are taken as mesenteries, the skeletal dissepiments, sclerosepta, as septa. 



We start, therefore, from a stage with 6 pairs of mesenteries arranged typically, but 

 with the lateral endocoels larger or at least as large as the exocoels. In each of the 6 

 endocoels a septum has been formed (text-fig. 4a). In the next stage the development 

 of mesenteries of the second order takes place in the same way as in the Endocoelactidae, 

 i.e. in the lateral endocoels, 4 pairs of the second order with the longitudinal muscles 

 turned outwards. Each of these mesenteries of the second order forms a new pair with 

 neighbouring mesenteries of the first order. In these new endocoels 4 septa are formed 

 (text-fig. 4b). 



Owing to this arrangement of the mesenteries and their occurrence only in the 

 lateral endocoels, 4 zones of development have arisen instead of the 6 found in the 

 exocoels of a normal Madreporaria. These zones of development lie one in each 

 quadrant of the animal. This results in an asymmetrical development of the 

 mesenteries, together with an irregular growth of the walls, due to the fact that the 

 animal is generally attached along the one side of the goblet, or at least has been so 

 once. The consequence is now that in each quadrant of the dorsal side of the animal, 

 i.e. the side turned away from the siphonoglyphe, a complete suppression of the 

 mesenteries of the next order takes place, while the development in the ventral part is 

 continued. In the ventral endocoels 4 pairs of mesenteries arise with the same arrange- 

 ment of the musculature as those of the second order. These mesenteries form new 

 pairs with adjacent mesenteries of the first and second order. In the 4 new endocoels 

 4 septa are formed (text-fig. 4c). The development is continued in this way with the 

 next order, with suppression of the mesenteries and septa in the dorsal endocoels of the 

 third order in each quadrant. At the end of the development, or at least at a late 

 stage, septa develop in the exocoels (text-fig. 4d). 



