for any month within a given year were weighted 

 in favor of certain days (dates) so that in com- 

 puting monthly means, all data for a given day 

 were averaged and monthly means based on the 

 daily means rather than on the total number 

 of observations. To indicate the density of cov- 

 erage and the variability to be expected in each 

 30-min quadrangle, the minimum and maximum 

 bottom-water temperatures, the number of 

 years' observations, and the number of day's 

 observations are tabulated by month and depth 

 zone in the Appendix. The few observations 

 made at depths greater than 250 m have not 

 been included in these tables. 



In all areas and at all depths the number of 

 observations were biased in favor of certain 

 months and years. To offset this bias, the 1940- 

 66 monthly mean values for each 30-min quad- 

 rangle area and for each depth zone within the 

 100-m isobath were plotted, and smooth curves 

 showing the seasonal cycle of temperature were 

 drawn as described by Fuglister (1947) and 

 Colton (1968a). In drawing these curves the 

 greatest weight was given to mean values rep- 

 resented in the greatest number of years. The 

 resulting curves for specific depth zones in ad- 

 jacent quadrangles were so closely similar that 

 the occasional inconsistencies in the data were 

 obvious. In general, the magnitude of the cor- 

 rection applied to the monthly mean value was 

 inversely proportional to the number of years 

 represented. 



Although there were appreciable monthly var- 

 iations in mean bottom-water temperatures at 

 depths greater than 100 m, there was no evi- 

 dence of a consistent seasonal temperature cycle. 

 In addition, the number of observations were so 

 few in any month in most depth zones and areas 

 that it was impossible to attach any significance 

 to the monthly mean temperature variations. 

 The analysis of long-term subsurface temper- 

 ature trends made by Colton (1968b) indicates 

 that these monthly variations are due in large 

 measure to the fact that the majority of obser- 

 vations in some months were made during pe- 

 riods of warming while in other months during 

 periods of cooling. To offset this sampling bias 

 we have based our estimates of monthly mean 

 bottom-water temperatures at depths greater 

 than 100 m on 3-month moving averages (Jan- 

 uary = December-February mean). 



TEMPERATURE CHARTS 



The monthly distribution of bottom-water 

 temperature is shown in Figures 2 - 13. In con- 

 structing these charts, contours were drawn on 

 a basis of corrected values read from 1940-66 

 mean seasonal temperature curves or deter- 

 mined from 3-month moving averages and en- 

 tered at the approximate geographic center of 

 each depth zone within 30-min quadrangle areas. 

 The boundaries of the depth zones were based on 

 the bathymetry given in U.S. Geological Survey, 

 Miscellaneous Geological Investigations Map, 

 1-451, Sheets 2 and 3. Isotherms were drawn 

 for each whole degree Centigrade as this ap- 

 peared to be an interval most useful to ecolo- 

 gists and one appropriate for displaying max- 

 imum resolution in the temperature structure 

 consistent with the quality of the data. In most 

 cases, the isotherms were drawn directly to the 

 data, but in some instances when an isolated 

 temperature value, based on a limited number 

 of observations, was unsupported by data in ad- 

 joining quadrangles, some smoothing and inter- 

 polation were necessary. Isotherms drawn on 

 the basis of limited data are indicated by a 

 dashed line. Relatively few observations were 

 made in water shoaler than 20 m and deeper 

 than 250 m, so that for the most part the iso- 

 therm lines were terminated at these depth 

 contours. 



Most useful statistics for delineating faunal 

 boundaries are the seasonal and annual ranges 

 of temperature in various geographical areas. 

 In general, the magnitude of the range of bot- 

 tom-water temperature decreases and the timing 

 of the seasonal temperature extremes occurs 

 later with increasing depth. The maximum and 

 minimum observed bottom-water temperatures 

 for the period of record are listed by month, area, 

 and depth zone in the Appendix. We had ini- 

 tially planned to present charts showing the 

 long-term monthly maximum and minimum bot- 

 tom-water temperature distribution, but the data 

 proved inadequate for this purpose. There were 

 a sufficient number of observations, however, 

 to construct valid distribution charts of the 

 long-term annual maximum and minimum bot- 

 tom-water temperatures. These charts are 

 shown in Figures 14 and 15 and are based on 

 the maximum and minimum observed temper- 



