1034 MISS R. HARRISON AND PROF. S. J. HICKSON ON 



conversely to the order in which they appear in ontogeny. But 

 it must be remembered that the stem of the Y is also the 

 representative of the secondary septum according to either the 

 popular or Pourtales' nomenclature ; and these septa still exist 

 in their normal position midway between the primary septa, 

 and project upwards as pali, or paliform lobes : in the lower 

 portion of the colony, before the secondary septa have begun 

 to branch, the costse corresponding to them are also nnbranched 

 and extend to the base. A comparison of the diagrams A, B, 

 in text-fig. 219, illustrates this point. 



Diagram A represents a plan of the costse and B a plan of the 

 septa ; in A a line "«a" has been drawn across the lowest part 

 of the corallum, and in B a corresponding circle " an" has been 

 drawn round the inner part of the septal plan; in both, the 

 costse and septa respectively of the first and second cycles are 

 present; presumably the coral must at one time have passed 

 through a sttige when these cycles only were present. In 

 T. fenestratus this is the arrangement in the adult. Again, 

 another line "&6" is drawn across A at a higher level and a 

 corresponding circle " 6 6 " on B ; in both, the costal and septa of 

 three cycles are present. When the secondary septa branched 

 peripherally, new costje arose in connexion with these branches, 

 and the original secondary eostfe remained in connexion with 

 the inner unbranched part of the secondary septa which persist 

 as pali or paliform lobes. Bourne [6, text-fig. I. 2j has given 

 a comprehensive diagram illustrating the relation of septa and 

 pali according to Pourttdes' principle : diagram B illustrates the 

 arrangement in T. zelandice-, which is entirely in agi-eement 

 w ith it. 



The fact that the costae of the fourth cycle are of different lengths 

 (see text-fig. 219 A, iv) is in accordance with Duerden's [11] 

 account of the septal sequence of Sklerastro&a radians, in Avhich 

 he states that new mesenterial pairs appear in some interseptal 

 chambers, before those in others of the same cycle. 



Agelecyathus persicus Duncan [14]. 



Cora.llites rising from an encrusting base, expanding slightly 

 from base to calice. Calice elliptical, depressed at one end of 

 the long axis. Height of corallum from 5-10 mm. at depressed 

 end of the long axis, and from 15-20 mm. at the opposite end. 

 Diameter of calice 9 x 7-5 mm. to 11x9 mm. The shape of the 

 corallites sTiggests that the colony was fixed on a vertical sur- 

 face, and the inequality of height in the two extremities of the 

 long axis is a response to an effort on the part, of the indi- 

 vidual corallites to assume a vertical position. Costse in four 

 complete cycles, distinct from base upwards, slightly exsert and 

 finely granular. Septa in six systems of four cycles of which 

 the last is incomplete, although all the costal of the fourth 

 cycle are present ; septa of the first two cycles more exsert 

 than those of the third cycle ; all three cycles reach the columella ; 



