OCEAN TEMPEEATUEBS. 173 



Eeef, at the former station occurring between August 1 and September 20, and at the latter be- 

 tween October 20 and December 8. 



EiGHTY-FiYE DEGEEKS.— The surfacc Waters seldom reach a temperature of 85° excepting for 

 short periods at the extreme south, and it has been impossible to construct isotherms for more 

 than one year at Martin's Industry and the Florida Eeef stations. Temperatures of 85° and over 

 were recorded between the 8th and last of July, 1881, at Eattlesnake Shoal, Fowey Eocks, and 

 the Tortugas; between the first part of July and the last of August, 1883, at Carysfort Eeef and 

 Fowey Eocks ; and in 1885, from August 2 to 2G, at Martin's Industry, from July 29 to September 

 28, at Fowey Eocks, and from July 9 to September 25, at Carysfort Eeef. 



Ninety degeees.— A surface temperature of 90° was occasionally recorded at some of the 

 extreme southern stations, but never for more than a day or two at a time. 



MEAN ISOTHERMS BASED UPON FIVE YEARS' OBSERVATIONS. 



In constructing chart No. 31, it was impossible, in all cases, to obtain ihe reductions of five 

 years' observations, on account of occasional breaks in the records, but the number of years 

 plotted is never less than three and generally more than four. The exact number in each instance 

 may be determined by reference to the annual isothermal charts (Nos. 26-30). 



On this chart the surface isotherms are represented as being much more regular and more 

 uniformly distributed with reference to time than on any of the yearly charts. On the left-hand 

 side of the chart, north of Body's Island, there is a wide area of low temperatures, bounded by the 

 isotherms of 40°. From Winter Quarter Shoal to Fire Island, inclusive, this area represents a 

 period of about seventy-five to one hundred days, being shortest at Five-Fathom Bank, and 

 longest at Sandy Hook ; from Block Island northward the length of this period varies from one 

 hundred to one hundred and thirty-five days, beiug shortest at Brenton's Eeef and longest at Boon 

 Island. A narrower space of maximum temperatures extends vertically through the center of the 

 chart, occupying principally the month of August, which is not crossed by any of the isotherms, 

 although a few of them extend a short distance into it. The length of time elapsiug between 

 successive isotherms is generally from about twelve to twenty-four days, seldom less but often more. 

 The isotherms are more numerous and follow one another more rapidly at the intermediate stations 

 of the series than at the northern and southern stations ; toward the north and south they diverge 

 somewhat and become more widely separated. This naturally results from the fact that at the 

 intermediate stations there is a much greater range of temperature (above a minimum of 40°) than 

 at the northern and southern ones. In the eastern part of the Gulf of Maine there are only three 

 continuous isotherms of the value of those plotted (40° to 50°, inclusive); in the western part four 

 such isotherms (40° to 55°, inclusive). On the southern coast of New England there are five 

 isotherms at Nantucket New South Shoal (40° to 60°, inclusive), and six from Vineyard Sound to 

 Block Island (40° to 65°, inclusive) ; seven extend from Fire Island to Body's Island (40° to 70°, 

 inclusive) ; four occur on the coast of South Carolina (55° to 70°, inclusive), and two at the Florida 

 Eeefe (75° to 80°, inclusive, and sometimes 85°). 



North of Body's Island the isotherms of both series extend in a slightly oblique direction treud- 

 _ ing inward toward the north. At Cape Hatteras they bend abruptly, and the same isotherms 

 appear much earlier in the year to the south of that important cape. 



At Frying Pan Shoals, the first reliable station south of Hatteras, the isotherm of 55° appears 

 about ninety days earlier than at Winter Quarter Shoal] the isotherm of 60° about sixty days 

 earlier ; the isotherm of 65° about fifty days earlier ; and the isotherm of 70° also about fifty days 

 earlier. 



