1898.] OP MILLEPOEA PROM FUNAFUTI AND BOTUMA. 831 



tabulae in each pore is about 17. The specimen appears to have 

 been in a very sickly condition when taken. More than | of one 

 face of it is dead coral, and the other face is considerably attacked 

 by Algae. Nearly the whole of the " live " surface is pitted with 

 Pyrgoma milleporce. In one place I counted no less than 13 young 

 cirri pedes in an area 1 x 1| centimetres. 



On the surface of a fragment which was probably broken off 

 this specimen there may be seen several ampullae. 



The specimens of Millepora collected by Mr. Grardiner in 

 Eotuma are of two kinds. They were found only in the boat- 

 channel, there being none on the reef. One of these consists of 

 coralla of light texture, of branching habit, similar to that usually 

 considered characteristic of Millepora alcicornis. The branches are 

 disposed in a single plane and freely anastomose, their average 

 thickness being about 5 mm. They are free from parasitic 

 cirripedes and show on some of the branches numerous ampullae. 



The other kind consists of very hard dense coralla, partly or 

 wholly encrusting dead coral, but as the free edges rise into plates 

 with crested borders they correspond most closely with the species 

 M. plicaia. Hence they may be considered under the term facies 

 "plicata." The thickness of the live corallum is rarely more than 

 3 mm. from each surface, and its great hardness affects the manner 

 of its fi'acture in such a remarkable way that great difficulties 

 present themselves when an attempt is made to count the tabulae 

 in each pore. From the (small) number of pores I have been able 

 to examine, I arrive at the conclusion that there cannot be on an 

 average more than five tabulae in each. 



One of these specimens shows the scars of numerous ampullae. 



The surface of all these forms from Eotuma is remarkably clean 

 and free from parasites of all kinds. 



The remarkable hardness of the corallum makes the pores very 

 apparent, and gives them the appearance of being much larger 

 than they really are. On first handling the specimen I thought 

 the pores were the largest I had seen, but on measuring the 

 diameters of 12 gastropores on one face of a specimen I found the 

 average to be only 0*27 mm., and on the opposite face the average 

 of five or six which I measured was less than 0*2 mm. These 

 figures show how deceptive estimates of size may be which are 

 made by unassisted vision. The pores of the facies "plicata" 

 from Eotuma are actually smaller than those of the facies " ramosa " 

 from Funafuti, and yet they have very decidedly the appearance 

 of being larger. 



The remarkable difference in size between the gastropores on 

 one face of the corallum and on the other which is recorded above 

 is by no means exceptional. In nearly every case in which I have 

 compared the average diameter of 12 gastropores from one part 

 of a corallum with an average of 12 on another I have found a 

 certain difference. It is probably to be accounted for by the 

 difference in food-supply, fresh water, or other external conditions 



