I6G FISHING-GEOU^'DS OF NORTH AMERICA. 



the raninlaiul, the two latter close to the oOfathoiu carve. Petit Manan Islaud, like Seguiii, lies 

 but a short distance ofif the mainland, and is the most northern and eastern station of the series. 



COMPARISON OF THE STATIONS WITH RESPECT TO TEJIPERATUKE. 



As explained elsewhere, the stations do not all afford temperature observations of equal value 

 on account of differences in the nature of their surroundings. Many of the stations included in 

 this report do not, therefore, furnish correct data with respect to the open waters of the coast, but 

 the character of the surface observations may be more or less accurately determined by a study of 

 their relations to the air temperatures. The light ships, being all located off shore in depths 

 exceeding 5 fathoms, are naturally best adapted for the taking of ocean temperatures, while nest 

 in order of excellence, as a rule, are the light-houses situated on small islands and reefs, more or 

 less distant from the mainland. The records for January and February at nearly all the stations 

 north of Cape Hatteras have not been used, on account of the manifest errors of observation some- 

 times made during extremely cold weather, by not reading the thermometer quickly enough after 

 it has been withdrawn from the water. These errors do not appear to extend much into either 

 December or March, although at some stations the records for those months may show too low a 

 range of temperature by a very small amount. However, the winter surface temperatures are not 

 of much importance in connection with any fishery problem north of Cape Hatteras. 



At the extreme south we recognize a group of stations which differ from all the others in the 

 conditions of temperature. It includes only the three light-houses of the Florida Reefs, bordering 

 the Gulf Stream. At Carysfort Reef and Fowey Rocks the curves of surface temperature are more 

 regular than at the Tortugas and correspond less closely with the air curves, indicating fewer local 

 influences or more open exposures at the places of observation. The three succeeding light ships, 

 Martin's Industry, Rattlesnake Shoal, and Frying Pan Shoals, afford more or less uniform results, 

 the plottings forming much more pronounced curves than at the Florida Reefs ; the surface curves 

 are most regular at the first mentioned light-ship. Cape Lookout is a shore station at which the 

 irregularities in the air curves are almost exactly repeated in the .snrfiice curves, although the 

 maximum surface tempeiature is no higher than atFrying Pan Shoals light-sliip. At Body's Island, 

 another station on the mainland, both the air and surface curves indicate extreme fluctuations in 

 temperature, which are almost precisely alike for both the air and surface. The surface lines at 

 Winter Quarter Shoal and Five-Fathom Bank light-sliiiis correspond closely in their general cur- 

 vature, and also in many of their details. At Absecou Inlet, the third shore station, the water 

 curves are more regular than at either Cape Lookout or Body's Island, but the maximum temper- 

 ature is the same for both the air and water. The surface curves differ considerably from those of 

 the air at Sandy Hook lightship, and at Fire Islaud and Block Island the results are much more 

 satisfactory than at any of the other shore stations, in both cases the maximum surface tempera- 

 tures being about 8o..j lower than the maximum air temperatures. At the three succeeding light- 

 ships, Brentou's Reef, Vineyard Sound, and Nantucket New South Shoal, the surface curves are 

 all comparatively regular; but at Pollock Rip lightship, they present many irregularities which 

 do not, in all cases, correspond with those of the air. The fluctuations of temperature are still 

 greater at Thatcher's Island, wliere the observations were taken in a sheltered position in .shallow 

 water. Of the islands in the northern part of the Gulf of Maine, the most regular and uniform 

 series of surface curves are presented by Matinicus Rock, Seguin Island affording the next best 

 series in that respect. At the three other stations the surface curves are less regular, especially 

 from 1881 to 1883, inclusive, but the irregularities do not appear to be due, in most cases, to atmos- 

 pheric influence. 



o 



