472 Prof. J. E. Diiculeii on the 



indicate very iiieogre protoplasmic contents. The nuclei are 

 fmall, irregular in shape, and comparatively lew in number, 

 and are aggregated towards the margin, where a few zooxan- 

 ihellai also occur similar to others distributed throughout the 

 inner layer. The tissue, as a whole, bears the closest 

 j-eseniblance to the endoderin found in the dee|)er, more 

 internal regions of most coral species. I have elsewhere 

 described this gradual hypertrophy which the endoderm of 

 coral l)olyps undergoes in passing from the upper to the 

 lower ])arts of the polypal cavity *. In the deeper regions 

 of the {'olyp the layer is nearly always greatly thickened and 

 vacuolated, with few protoplasmic contents and small nuclei, 

 'i'he combination of characters indicates a tissue in a non- 

 active condition, not one concerned in the growth or digestive 

 processes of the polyp; moreover, the modification is never 

 found over regions of the polyp where the skeleton is in 

 process of active formation, these being in the upper growing 

 paits of the pol} p and corallum f. 



Enquiry may now be made as to the origin and function 

 cf the downgrowths. As regards the former there can be no 

 question of the direct relationship of the downgrowths with 

 the tentacles, from the fact that they occupy the discal position 

 cf these orgf ns and replace them. In Stylojyhora they present 

 altogether the characters of introverted tentacles, having 

 endodeini on the outside, ectoderm on the inside, and a narrow 

 lun.en con;municating with the exterior. The phenomenon 

 of tentacular introversion is of very general occurrence in 

 living j\Iadreporarian polyps; moreover, it is the usual 

 condition of retracted preserved polyps. I have observed it 

 in numerous West Indian and PaciHc corals, but no satis- 

 factory explanation of the process has yet been forthcoming. 

 Introversion and extrusion of single tentacles have been 

 observed to be repeated several times in close succession, and 

 may even take place wliile the rest of the tentacles and the 

 polyp as a whole remain fully expanded. Erom the manner 

 in which it })roceeds it is evidently connected with the 

 internal circulation of the polypal fluid, a diminished pressure 

 in any mesenterial chamber leading to an introversion of the 

 tentacle communicating with it, and an increase in the pressure 



* ''West Indian Wadreporaiian Poh-ps/'Nat. Acad. Sciences, Mem. 7, 

 \ol. viii. 1902, p. 480. 



t FoclUupora is exceptional among corals in tbat the internal endo- 

 derm is no thicker than that over the n st of the polyp. On tlie other 

 hand, the tkeletogtnic ectodenn, even where apparently non-active, is 

 much Iroader than in mo.«t coral.«, though cellular distinctions are not 

 marked (tigs. 1 and :.'). 



