MATTHAI— RECENT COLONIAL ASTR^ID^ 17 



Gardiner has recorded the general occurrence of nematocysts in the endoderm of the 

 polyps of Flabellum on each side of the upper top ends of the septa. These undoubtedly 

 belong to III, which the following considerations render it probable that they may have 

 been at first formed in the calicoblastic layer, their subsequent presence in the endoderm 

 being perhaps due to displacement or migration : {a) nematocysts of this type are present 

 in the calicoblastic layer — compare type II in the calicoblast of Mycedium okeni Ed. 

 and H. — usually towards the top ends of the septa, some of them being entire with sac, 

 protoplasmic core and coiled thread and lying parallel to the septa with the calicoblastic 

 layer raised over them or in an oblique position partly in the calicoblast and partly in the 

 endoderm ; (6) in the endoderm no entire nematocysts are present, the wall being, in 

 every case, absent and the coiled thread usually broken ; (c) they have no definite 

 position, but lie loose in the endoderm ; {d) the nematocysts in the calicoblastic layer are 

 well stained, whereas in the endoderm they have taken very little stain, probably because 

 of subsequent degeneration. 



Gardiner makes no record of such nematocysts in the mesenterial filaments of 

 Flabellum, but oval, pink areas with a suggestion of coiled threads can be seen in the 

 convolutions of the filaments*. 



Nervous system (PL 1, fig. 3, PI. 2, fig. 24, PI. 5, figs. 47, 52, PL 6, fig. 66). Nerve 

 cells and fibres t as figured by the Hertwigs for the Actinians have not been met with in 

 my sections. The ectoderm just above the mesogleea is more granular than usual, being 

 the only indication of a nervous stratum, perhaps representing their "Nervenfaserschicht" 

 and von Heider's " Interbasalsubstanz." This is seen in the edge-zone, oral-disc, 

 tentacles, ectodermal ridges of the stomodseum and the mesenterial filaments, and best 

 of all in the tentacles, perhaps because of their bundles of nematocysts. 



Endodermal musculature. This constitutes the main musculature of the polyp ; it 

 is present in two regions, viz., oral-disc and mesenteries. 



Attached to the lower surface of the mesogisea of the oral-disc (PL 5, figs. 51, 52) is 

 a thin layer of muscle-fibres disposed circularly, but much less developed than in Actinians 

 and without any mesoglaeal pleats which are so prominent in many of the latter {vide 

 R. Hertwig, Haddon and van Beneden). Around the outer margin of the bases of the 

 outer cycle of tentacles the muscular layer is somewhat thickened to form a sphincter, the 

 so-called Rotteken's muscle ; from the appearance of the spirit specimens it may be inferred 

 that its action is to reduce the opening of the peristome by drawing its lateral wall over 

 the oral-disc and hence over the tentacles, which consequently are bent over the mouth. 

 Faurot (23) does not regard the sphincter as having any retractile action on the tentacles 



* In Coenopsamviia Gardiner's figs. 7 and 8 are type I ; 9 and 10 are appearances in the protoplasm ; 14 is 

 a combination of types II and III; 15 and 16 are probably type III; 17 — 19 are type II but the nucleus is 

 an optical appearance. 



t Havet (61) employing better methods has been able to distinguish both sensory and motor nerve cells 

 and fibres in the ectoderm and endoderm of an Actinian Metridium dianthus, and has given excellent figures 

 illustrating their arrangement and distribution. Van Beneden could not make out any ganglion cells in his 

 sections of larval Cerianthidse, although a distinct "assise nerveuse" was present in both the ectoderm and 

 endoderm at the base of the "assise ^pithelioide " and just above the "assise musculaire." Faurot could not 

 find such a nervous layer in the endoderm of the Actinians he studied. 



SECOND SERIES— ZOOLOGY, VOL. XVII. 3 



