of some Corals from Fiji. 219 



branches without twigs, the position of growth may be nearly 

 accurately fixed on the dry specimen. Massive species also show 

 well-defined characters, but these vary with each genus, and can 

 only be learnt by long experience on the reefs themselves. 



Finally I made a rough estimate of the number of polyps 

 on some of these coralla, but the method is of little use, as the 

 polyps of different species are not all of the same size, nor have they 

 necessarily the same method of budding, nor indeed any connection 

 with the volume of the skeleton. It is interesting however to 

 note that, while No. 5 gives a sheet almost twice as thick as 

 Nos. 1 and 3, it actually has only about half as many polyps, the 

 calculated numbers of 1, 3 and 5 being respectively 25470, 23616 

 and 12672. Assuming an actual age of 1030 days and a regular 

 geometrical progression of 2, periods of 71, 70, and 76 days would 

 have elapsed in Nos. 1, 3 and 5 respectively between each set as 

 it were of buds 1 . Examined in this way, the numbers of the 

 polyps do not seem very large, considering the peculiarly favour- 

 able situation of the colonies. It is again suggested that the 

 actual time of growth must have been much less than the observed 

 time. 



1 I am indebted to Mr J. F. Cameron, Fellow of Gonville and Caius College, for 

 this calculation. 



