34 Temperature of the Ocean. 



unprotected thermometers previously in use sometimes exceeded 

 io° C. for the same depth ; and a comparison of the data 

 furnished by the Miller-Casella thermometer with the corre- 

 sponding temperature curves shows that the mean error of all 

 the deep-sea observations made on board H.M.S. "Challenger," 

 not much under 10,000, is probably less than o c -5 C. The 

 thermometers, before being sent out, are subjected to pressures 

 varying from one to four tons to the square inch in a hydraulic 

 press especially designed for this purpose by Mr. Casella, and 

 the amount of error ascertained for each instrument. When in 

 actual use, they are enclosed in a copper cylinder perforated 

 at both ends to allow free ingress and egress to the water. 

 Several thermometers may be attached to the same sounding- 

 line whilst it is being paid out, at distances of five, ten, twenty- 

 five, fifty, or one hundred fathoms as required ; and it is found 

 that an immersion of from five to ten minutes is sufficient to 

 secure the desired record of temperature. 



Serial Soundings and Temperature Curves. — In order to 

 ascertain the distribution of temperature at any selected station 

 upon the ocean, it is necessary to obtain as many observations 

 at various depths between the surface and the bottom as circum- 

 stances will permit. The amount of time absorbed in hauling 

 in the sounding-line from great depths, even with the help of a 

 steam-engine, as well as the uncertain state of the weather, 

 which may interrupt the operations at any moment, necessarily 

 limit the number of observations which can be taken in the 

 course of a stoppage on the high seas. However, the first 

 experiments made of this kind establish the fact that, beyond a 

 depth of from 1000 to 1500 fathoms, the temperature of the 

 water decreases very slowly, often not more than o°.i C. in 100 

 fathoms ; and that, beyond these depths, observations taken at 

 intervals of 100, 200, or 250 fathoms are sufficient for all 

 purposes. A series of observations at everv 10 fathoms from 



