CORALS FROM THE PERSIAN GULF. 



1033 



paper, and the occurrence of the same species in localities so far 

 apart is remarkable. 



I had hoped that the sections might throw some further 

 light on the question of septal sequence, but unfortunately the 

 lower fifth of the coral has been filled up solidly by a secondary 

 deposit of calcareous matter, and the interseptal chambers do 

 not begin to appear distinctly until a level is reached where 

 twelve septa and twenty-four costse are already present. If it 



Text-fiij. 219. 



'• II. III. lY 



A. DiaTfammxtic plan of the co?t;B of Trem'itoti'ocJms selanii:e, showing the 



difference in length of the costas of the fourth cycle, i, ii, iii, iv, costaj of the 

 first, second, third, and fourth cycles. 



B. Diagrammatic plan of the septa of Trematotroohus zelawUm, representing the 



exsert peripheral portion of the primary septa, and the exsert pali of the 

 secondary septa, i, ii, iii, septa of the first, second, and third cycles ; ex.s., exsert 

 portion of the septa of the first cycle ; p., pali. 



may be taken that the costae correspond to the septa and that the 

 base of a coral represents the earliest formation, it might be 

 objected that Pourtal^s' law for septal sequence has not been 

 followed in this case ; for the costfe which extend to the base, 

 twelve in number, are those corresponding to the first and 

 third (according to Pourtal6s [35]) cycles of se^Dta. That is to 

 say, if the secondary septa split peripherally in a Y-shape, and 

 a tertiary septum grew up between the arms of the Y, the 

 septa generally spoken of as secondary and tertiary are named 



