42 



Papers from the Department of Marine Biology. 



than would be made in the normal thickening of the branches. Thin 

 laminar expansions grow out from the sides of the branches in a plane 

 parallel to that of the crab's respiratory current. These laminae do not, 

 however, represent continuous growth, but are formed by the coales- 

 cence of parallel twigs which can be easily distinguished at the edge 

 of lamina. This is perhaps best illustrated by a specimen in the Cam- 

 bridge University Museum of Zoology, which is shown in figure SE. 

 The numerous gaps which occur in the walls of the gall are due to the 

 incomplete fusion of the component twigs. 



- o::-'V: 



SeP",-; 



<3A^l 



M-^^3M.-o o- ,es. 

 -*:^.o : -.V ; -QVO / 9 e - 



-.. -To o o '--'. 



;?^ox * 



0. 



.;:^ 



^:^^f^i^?-^?^ 



':-. . :': o.:^.o ,o. o o ' ^ 



^g^^^5f^^ 



CO. 



B 



FIG. 4. Surface view of colony of Pocillopora c&spitosa. 



A outside and B inside a gall. X20. After prolonged decalcification, preparations stained with 



borax-carmine. 

 In A the polyps arc regularly developed, the tentacles and mesenteries are formed of thick, deeply 



staining tissue and the septa are of normal number and symmetrically arranged. 

 In B the polyps are stunted and irregular, the mesenteries and tentacles not well developed and 



the septa often placed asymmetrically and sometimes short of the typical number. 



co., coenenchyme, ; oc., oesophagus; mes., darkly staining tissue, mesenteries and tentacles 



indistinguishable; sep., septa. 



The initial result of the interaction of the crab and the coral is the 

 production of a small initial chamber as in Pocillopora. Afterwards 

 an upper chamber is formed, large enough to contain the mature 

 female. The complete gall is not quite so regular as that in Pocillopora, 

 but it has essentially the same structure. 



At first sight the interference with the normal habit of branching 

 involved in gall formation seems to be much greater in Seriatopora 

 than in Pocillopora. In the latter case there is an apparent broadening 

 of existing branches, a stimulation of lateral growth as compared with 

 apical growth. In the former, however, there is at the point of settle- 

 ment a production of numbers of tiny lateral branches which, together 

 with the main branches from which they spring, form the walls of the 

 gall a basketwork the meshes of which are filled in by subsequent 

 growth. The course of the main branches can always be traced, even 



