82 PERCY SLADEN TRUST EXPEDITION 



protoplasm between which mucous vacuoles are frequently present ; m the lower half 

 of the layer the protoplasm is clearer with scattered round nuclei. 



Filaments are present on all the mesenteries. Nematocysts I are quite common in 

 the straight regions of the filaments of the principal mesenteries, less frequent in those 

 of the subsidiary mesenteries ; in the coils numerous nematocysts II and III are present, 

 the former with the dark-stained axis slightly bent and the latter with the thread partly 

 extruded ; type I is scarce. The endoderm has a slaty-blue colour, conspicuous in 

 the mesenteries ; in the oral-disc it is thinner than the overlying ectoderm ; the circular 

 layer of endodermal muscle-fibres is well seen. The tentacular endoderm is less than half 

 the thickness of the sub-terminal batteries, with comparatively few algae. The mesenterial 

 endoderm is only slightly swollen behind the filaments ; incomplete nematocysts II are 

 occasionally met with in it. Algae are comparatively rare, occurring mostly on 

 the entocoelic side in the outer pleatal region and on the exocoelic side in the inner 

 pleatal region. To the inner exocoelic pleats thick muscle-bands are attached which 

 make up for the somewhat weak musculature on the entocoelic pleats. Below the 

 stomodseal region of the polyp, the exocoelic pleats do not extend, but the entocoelic pleats 

 cover almost the entire widtii of the mesenteries. One, two or three longitudinal rows of 

 somewhat small ova were present in all the polyps, in some of the principal and subsidiary 

 mesenteries. 



Number of polyps examined, fourteen, taken from ten different specimens of varying 

 skeletal facies, all from Dongonab, Red Sea. 



B. Corallum. The exact position of the specimen from Minikoi, measuring 

 12'5xll"5x6 cm., which Gardiner referred to Favia versipora (Ehrb.), is doubtful 

 (PI. 21, fig. 4). As pointed out by him it comes nearest to the typical form of 

 Klunzinger's Favia ehrenbergi (PL III, fig. 7). Its corallites are polygonal, often elongated 

 for fission, the single calicos being quite small, about 4 mm. wide and as much deep (the 

 diameters of the longest calyx with signs of fusion into three being 10 and 3"6 mm.). The 

 inter-calicinal walls are flat above, about 2 mm. thick. The septa are few in number, about 

 20, of which 10 — 12 meet the columella; they are thick, vertical, with rough sides and 

 4 or 5 bluntly-pointed horizontal teeth on their edges. The septa of adjacent corallites 

 are usually continuous over the walls, being about "5 mm. exsert; the exsert ends are 

 sometimes connected along the middle of the walls by thin longitudinal ridges. The 

 columellee are formed of dense trabeculee, often quite compact and ^ — ^ the width of the 

 calicos. Fission is sub-equal. The specimen also approaches PI. 21, fig. 3, in appearance, 

 but the latter has larger corallites and thicker walls. 



Marenzeller has already pointed out Klunzinger's error in regarding Madrepora 

 favus, Forsk. as a species of Goniastrea. Five of Forskal's examples of this species are 

 at present in the Copenhagen Museum, one of which (21 X 17 X 15 cm.) comes near Milne 

 Edwards and Haime's type of Favia hertholleti. The remaining four (PL 22, figs. 1 — 4) 

 are skeletal varieties of the present species; one of these (Ilxl0'5x9 cm., fig. 4) has 

 thick perithecal regions, distorted corallites, approaching the condition in Klunzinger's 

 type of Favia ehrenbergi, var. laticollis. Madrepora cavernosa, Forsk. is represented by 

 a single specimen (17 x 17 x 11*5 cm., PL 22, fig. 5), which retains the normal charac- 



