SECT. 2] 



LARGE-SCALE INTERACTIONS 



153 



For application of the results to the Caribbean ellipse via equation (1), it is 

 now only necessary to multiply B by the surface area of 2.19 x 10^^ cm'-. 



One of the most intriguing parts of this Caribbean study by Colon was his 

 computation of the storage S to put in (1) directly from oceanographic measure- 

 ments. Later comparisons of the results for Qs and Qe with those of the transfer 

 formulas and with atmospheric heat and \\ater requirements permit a more 

 critical test of the storage determination than has been possible in previous 

 estimations (cf. Pattullo, 1957). To evaluate S, first the annual temperature 

 cycle and its distribution with depth to the vanishing point is required. For this 

 purpose a tabulation of about 8000 monthly temperature soundings, averaged 

 for one-degree squares, was compiled from bathythermograph data at the 

 Woods Hole Oceanographic Institution. In constructing the ellipse-averaged 

 temperature- depth distribution for each month shown in Fig. 27, a small but 



JFMAMJJASOND 



Fig. 27. Seasonal march of water temperature in Caribbean Sea from surface down to level 

 (in meters) where seasonal cycle is assvimed to vanish. Curves at depth obtained from 

 bathythermograph data of Woods Hole Oceanographic Institution. Amplitudes 

 adjusted slightly to fit surface curve adopted from Fuglister's (1947) charts. (After 

 Colon, 1960. Fig. 10.) 



iinjiortant adjustment was made, presumably to correct for slightly un- 

 representative sampling in summer when the data coverage was sparsest : the 

 annual surface temperature range would have been 5.9°F from the bathy- 

 thermograph sample ; it was reduced to 4.6°F to agree with climatic data for 

 the region (Fuglister, 1947; Atlas of Climatic Charts of the Oceans, U.S. 

 Weather Bureau, 1938) and the amplitudes at depth were adjusted propor- 

 tionally. The annual range is essentially undiminished at 15 m, reduced to 87% 

 at 30 m, to 76% at 45 m and to 26% at 75 m. At 90 m, the seasonal cycle was 

 assumed to have vanished ; below that it is very small and, if anything, appears 

 to change phase. 



