of bottom-water temperatures. Such a compen- 

 dium would serve as a norm to which individual 

 cruise data could be compared. It was not until 

 recently, however, that sufficient data were 

 available to construct truly representative con- 

 tour charts in which the isotherm intervals were 

 sufficiently small to accurately delineate homo- 

 genous regions. 



To meet this need we present in this paper, 

 charts of monthly average bottom-water tem- 

 peratures and of long-term annual minimum and 

 maximum bottom-water temperatures in the 

 area of the continental shelf between Nova Sco- 

 tia and New Jersey (Fig. 1). We hope that 

 these charts will prove useful in defining faunal 

 regions and lead to a better understanding of 

 the relation of temperature to the distribution 

 of benthic organisms. 



SOURCE DATA 



The long-term temperature data are for the 

 period 1940-66 and are from the file of bathy- 

 thermograph trace prints at the Woods Hole 

 Oceanographic Institution. This is the most 

 complete catalogue of temperature data for the 

 area and consists principally of observations 

 made by the Woods Hole Oceanographic Insti- 

 tution, Canadian Naval Research Establishment, 

 Fisheries Research Board of Canada, U.S. Coast 

 Guard, U.S. Coast and Geodetic Survey, and the 

 Bureau of Commercial Fisheries (now National 

 Marine Fisheries Service) . The year 1940 was 

 chosen as the starting point as this was the year 

 of the introduction of the bourdon-type bathy- 

 thermograph (Spilhaus, 1940) and the begin- 

 ning of a period of extensive temperature mea- 

 surement in the area. The bathythermographs 

 used throughout this ])eriod were calibrated in 

 various temperature and depth intervals. The 

 maximum depth limit of any bathythermograph 

 used extensively was 275 m, and this isobath 

 was used to delineate the oflFshore limit of the 

 area of coverage. 



EDITING AND ANALYSIS 



In order to insure a sufficient number of ob- 

 servations to make valid spatial and temporal 

 comparisons, it was necessary to treat the data 

 in terms of 30-min quadrangle areas (Fig. 1). 



Bottom-water temperatures over most of the 

 area are governed chiefly by depth, but also to 

 some extent by locality. Thus, to insure posi- 

 tional accuracy of the isotherms it was neces- 

 sary to subdivide the 30-min quadrangle areas 

 into depth zones and to detei'mine average tem- 

 perature values within these zones. The depth 

 limits of the zones used were determined in part 

 by the number of observations available, but also 

 by the nature of the seasonal temperature cycle 

 at specific depths. The majority of the obser- 

 vations were in depths shoaler than 100 m. Bot- 

 tom-waters at these depths are characterized by 

 pronounced seasonal temperature variations and 

 a complex and unstable temperature structure 

 which is controlled for the most part by solar 

 heating and wind stirring. At depths greater 

 than 100 m the bottom-water temperature struc- 

 ture is less variable and controlled for the most 

 part by advection. The nine depth zones des- 

 ignated were as follows: 1-20 m; 21-40 m; 

 41-60 m; 61-80 m; 81-100 m; 101-150 m; 151- 

 200 m; 201-250 m; and greater than 250 m. 



Bottom temperatures were read to the nearest 

 0.1°. The maiority of the temperature values 

 were in degrees Fahrenheit. All Celcius values 

 were converted to this scale for initial tabula- 

 tion. In editing the individual bathythermo- 

 graph traces, a comparison was made of bathy- 

 thermograph, echo sounder, and chart depths. 

 Observations were excluded in situations where 

 these soundings varied significantly or in which 

 it was obvious that there was a position error 

 or that the bathythermograph did not reach bot- 

 tom or a depth where the bottom-water temper- 

 ature could be accurately interpolated. After 

 the initial editing, the data included approxi- 

 mately 22,000 observations. 



The following data for each observation were 

 then entered on punch cards: latitude and 

 longitude in degrees and minutes, 30-min quad- 

 rangle area location coded as shown in the Ap- 

 pendix, bottom-water temperature in 0.1 °F, 

 depth in meters, and depth zone allocation num- 

 ber. A computer program was written to give 

 a listing by 30-min quadrangle area, month, and 

 depth zone of the number of observations, num- 

 ber of days' observations, number of years' ob- 

 servations, average bottom-water temperature, 

 and maximum and minimum bottom-water tem- 

 peratures in degrees Centigrade. Observations 



