SUMMARY AND CONCLUSIONS 365 



structure shown in figure 41 may with profit be contrasted wilh llic 

 typical trepostonie structure shown in figure 38 of the same ])late or in 

 figures 23 to 37. Agaiii in longitudinal sections of Merlin the cross- 

 partitions (diaphragms) have a large central perforation with a down- 

 ward-turning collar — iinlikr anything e^^cr noted in the t ii'puslomes. 

 The fact is that the structure of Merlia is about as different from that of 

 the trepostomes as anything could well be. Kirkpatricl< was completely 

 mislead by superficial resemblances. 



SUMMAKY AND CONCLUSIONS 



This paper deals with a number of morphological and hislological 

 characters of the Trepostomata which are either new or have hcix-tofore 

 been imperfectly understood. 



Cysts and cystiphragins. — More or less perfectly formed calcareous 

 cysts inclosing peculiar brown material are described and their relation 

 to cystiphragins explained. It is suggested that these structures are 

 developed in connection with successive degenerations and regenerations 

 of the polypides, and that the purpose of cystiphragms is the restriction 

 of intrazooecial space. 



Communication pores. — The histology of communication pores in the 

 Trepostomata and in the genus Heteropora is described, and it is shown 

 that not only are these structures probably universally present in the 

 Trepostomata, but that the pores have the same characteristics and ar- 

 rangement as in Heteropora. > 



hitrazocecial spines. — Certain extraordinary spines projecting into the 

 submature region of zooecia of a species of Nicholsonella are described 

 and their resemblance to the spines of Heteropora neozelanica, as figured 

 by Nicholson, is pointed out. 



\Yall structure. — The structure and histology of the walls of the 

 Trepostomata, as seen in longitudinal sections, is minutely described. It 

 is shown that the (ii\isions Integrata and Amalgamata, hascd on ihc 

 supposed presence or absence of a definite divisional plane in the center 

 of the wall, are o])eii to some question, and that the trepostonie wall was 

 probably single and cominoii to adjoining zooecia, as it is in the Cyclo- 

 stomata. The method of oi'igin and the varying arrangements of the 

 wall lamina' are descrihcil. aiid it is shown that the presence or absence 

 of a dai'k median line in the wall ile]K'nds to a large extent on Uie steep- 

 ness of pitch of (he lamina', which in turn depends on whether the grow- 

 ing edge of the wall is thin and sharp or blunt and smoothly rounded. 

 It is shown thai the .secondary deposits (or cingulum) are delinilelv re- 



