ANATOMY OF THE MADREPOKAEIAN CORAL FUNGJA. 295 



the base, and it is not unlikely that they may be formed only 

 when, by some accident, the coral has been overturned. 

 Examples of fission are rare, but I have in my possession a 

 dead corallum which is nearly divided into two separate 

 Fungise, and in which the axial fossae are already completely 

 separated from one another and form mouths excentrically 

 placed on peristomial discs inclined towards one another at a 

 wide angle. There is a similar specimen in the British 

 Museum. In the same collection there is a very good specimen 

 of a nurse-stock brought by H.M.S. Alert from the Seychelles, 

 found on March 5th, 1883. This is a young specimen from 

 which the first bud has not yet been detached, and the soft 

 tissues still extend down over the outside of the corallum to 

 the basal disc ; unfortunately the spirit in which it was con- 

 tained has been allowed to evaporate, and the soft tissues are 

 unfit for examination. But since it was found only two months 

 later than the date of my search at Diego Garcia and in the 

 same seas, it may be taken as an objection to my opinion 

 given above, that there is a special season of sexual followed 

 by asexual reproduction from a nurse-stock in Fungia. It is 

 quite possible that sexual reproduction may be very much 

 economised in these corals and is of rare occurrence, the 

 maintenance of the numbers of a species being ensured firstly 

 by the budding off of an indefinite number of forms from the 

 sexually produced nurse-stock, and secondly, by the simple 

 asexual processes of budding and fission above described. 

 The whole history of the reproduction of these forms is very 

 imperfectly understood, although it presents many problems 

 of the greatest interest. A naturalist travelling in coral-seas 

 should not fail to try and secure some specimens of the nurse- 

 stocks carefully preserved in spirit, as well as specimens of the 

 young free forms recently separated from the parent-stock. To 

 this should be added any observations that may be possible on 

 the relative frequency of the nurse-stocks, on the frequency of 

 budding or fission, and on the rate of growth. I was unable 

 to carry out an extended series of observations on the Fungiae 

 at Diego Garcia, for the knolls on which they were found lay 



