322 



PAPERS FROM THE DEPARTMENT OF MARINE BIOLOGY. 



and cooling endured by corals. The actual reef records show that reef 

 corals in nature do not withstand so much cooling as in the laboratory 

 experiments. The data at hand indicate that a well-developed fossil reef was 

 not subjected to an annual minimum below i8° C, and that the temperature 

 was so low at only rare intervals. The mean temperature for the coldest 

 months must have been not lower than about 21° C. (about 70 F.), as it 

 seems that a continuous bottom temperature of about 21 C. is the lowest 

 at which reef-forming corals may be expected to live. 



One of the needs in the further investigation of corals is more accurate 

 information on bottom temperature between 50 and 300 meters in coral- 

 reef areas. The data here presented are interesting and important, as they 

 constitute a beginning in the accumulation of such data. 



The information contained in the following tables has an important 

 bearing on many oceanographic problems, among which is the capacity of the 

 water to hold CaC0 3 in solution. This subject is discussed in preceding 

 papers, pages 267-268, 278, 287, 288, 315-317 of this volume. 



It is needless to say that this large body of information was only 

 assembled by me. I heartily thank all who have so generously assisted in 

 making it available for this paper. 



Summary of climatic data on southern Florida. 



The summary of climatic data on southern Florida as given in the 

 foregoing table was supplied by the U. S. Weather Bureau. 



The following weekly air-temperature records and the monthly averages 

 for Tortugas were contributed by Dr. Alfred G. Mayer. 



