70 CORAL AND ATOLLS 



to see, for they take place quite irregularly in different 

 colonies ; but no colony measured in the atoll was exempt 

 from them. In 100 days, from December 1, 1905, to March 

 10, 1906, a Madrepora hrsiuch. grew 3"5 centimetres; and 

 then from March 10 to July 24, a period of 136 days, no 

 addition was made to its growth, and the branch though 

 perfectly normal and healthy showed absolutely no signs of 

 activity. 



In one month, from December 31, 1905, to January 31, 

 1906, two Pontes masses showed no increase in their measure- 

 ments, and yet by July 31, one had added 9 centimetres to an 

 original circumference of 180, and the other 13 to an original 

 circumference of 219. 



The period of the greatest activity of the Madrc]}OTa was 

 the period of the most complete quiescence of the Poritcs and 

 vice versa; and yet all the growths were measured under 

 identical corditions, all lived in close proximity, and Avere 

 subjected to the same influences of tide and season. The 

 question of the cause of this waxing and waning of coral 

 growth is interesting, and one that doubtless has great in- 

 fluences on the building of dry land, for the prolonged 

 inactivity of any of the more important builders would 

 necessarily mean an alteration in the rate of land-formation. 

 It is obvious that any observations that only extend over brief 

 periods of time cannot be taken as strictly accurate records of 

 the rate of growth of corals, for the measurements may chance 

 to be made during a passive phase, or during an active phase 

 of the colony. 



In order to eliminate in some measure the influence of 

 this variability of growth, the observations were carried on for 

 as long as possible. The chances of the sea and of storms cut 

 many experiments short, but no case Avas used for obtaining 

 results in which the measured period of growth had not 

 exceeded 100 days. With the massive forms of growth this 

 method of measurement could not of course be adopted, and 

 direct measurements of the maximum circumference had to be 

 taken at intervals. When a groAvth of Poritcs has reached 



