CHAP, vii.] DEEP-SEA TEMPERATURES. 287 



this way imperceptible by any direct effect upon 

 navigation beyond the 45th parallel of north latitude, 

 a peculiarity which has produced and still produces 

 great misconceptions as to its real character. 



The mode of determining the surface temperature 

 of the ocean is sufficiently simple. A bucket is 

 let doAvn from the deck of the vessel, dashed about 

 for a little in the water to equalize the temperature, 

 and filled from a depth of a foot or so below the sur- 

 face. The temperature of the water in the bucket is 

 then taken by an ordinary thermometer, whose error 

 is known. A common thermometer of the Kew 

 Observatory pattern graduated to Fahrenheit degrees 

 can be read with a little practice to a quarter of a 

 degree, and a good- sized centigrade thermometer to 

 a tenth. Observations of surface- temperature are 

 usually made every two hours, the temperature of 

 the air being taken with each observation, and the 

 latitude and longitude noted at noon, or more fre- 

 quently by dead reckoning if required. 



Every observation of the surface-temperature of 

 the sea taken accurately and accompanied by an 

 equally exact note of the date, the geographical 

 position, and the temperature of the air, is of value. 

 The surface observations taken from H.M.S. ' Por- 

 cupine' during her dredging cruise, in the summer 

 of 1869, are given in Appendix A. 



The surface-temperature of the North Atlantic has 

 been the subject of almost an infinite number of such 

 observations, more or less accurate. Dr. Petermann, 

 in a valuable paper on the northern extension of 

 the Gulf-stream, reduces the means of more than a 

 hundred thousand of these, and deduces the scheme 



