CORALS FROM THE PERSIAN GULF. 1025 



alternating costse, paliform teeth exsert before all septal cycles 

 except the last, and an ill-developed papillose columella which 

 scarcely rises above the surface of the broad shallow central fossa. 



When Miss Poole stated thai; the forms in the Burmese 

 collection belonged to three different Types, two of which were 

 identical with Gardiner's Types I and II, she had not the advantage 

 of seeing the actual specimens, and was in error. All the 

 Burmese forms belong to Gardiner's type I, but they show great 

 variation, and in order to justify their inclusion with H. (equi- 

 costatus three types must be recognized. These types may be 

 called A, B and to avoid confusion, but it must be understood 

 that they correspond with Miss Poole's Types I, II & III. 



Type A = Type I as defined by Gardiner. There is normally a 

 single small aperture on the basal surface in addition to the 

 opening of the Aspidosiphon chamber. In a few, this additional 

 aperture is absent, whilst in others, two or at most three such 

 apertures occur. 



Type B is a much lower flattened form without a fossa; a 

 tendency for the costse to alternate ; four septal cycles, the 

 exsertness of the septa of the fourth cycle on either side of the 

 primaries is very marked, and the joining over the tertiaries 

 and secondaries gives a characteristic star-like appearance ; the 

 columella is less well-developed, it is a compact trabecular mass, 

 and not distinctly papilliform as in Type A ; the pores of the 

 endodermal canals are more numerous, and are not restricted to 

 the base, but are distributed irregularly round the lower part of 

 the corallum. 



Type C is a taller form, the base tends to equal the disk in 

 diameter, and the walls are nearly perpendicular ; a fifth cycle of 

 the septa is present ; there is a distinct fossa, and the centrally 

 depressed columella is a trabeculate mass as in the previous type; 

 the pores of the endodermal canals are disposed in a ring a little 

 below the calicular margin. 



In defining these three types, I have retained Gardiner's Type 1 

 and Miss Poole's Type in order to avoid confusion as far as 

 possible ; but it must be borne in mind that Type B as here 

 defined is very difi"erent from Gardiner's Type II. 



Practically the whole of the South African collection belongs to 

 Type A, the Burmese collection belongs to Types B and C ; the 

 specimens Miss Poole described as belonging to Type I should be 

 included in Type B. The Ceylon collection described by Professor 

 Bourne is somewhat intermediate between Types A and B. These 

 three types will embrace the species of Semper, Rehberg, and 

 Alcock*. 



H. parasiticus Semper [38]. Intermediate between Types A 



and B. 

 H. philippinensis Semper [38]. Two types. PI. xx. fio- 12 is 



Type B. PI. xx. figs. 13 & 14, Type 0. * 



* I have not considered S. sulcattis and H. lamellosus Verrill, and S. cochlea 

 Gmelin, as I have not been able to obtain first-hand reference to these species. 



