192 University of California Publication*. [Vol.6 



The volume of the bucket is about twenty liters, and since a 

 test showed that the temperature of the contents rose from 15.5° 

 to 15.75° when standing in the open air at 19.5°, in 20 minutes, 

 it was thought that deep-water temperatures could be determined 

 with but little error. However, when drawn up from a depth 

 of two hundred fathoms or more, the temperature at the top of 

 the bucket was repeatedly found to be as much as 2° or 3 D 

 warmer than at the bottom. The bottle being of bronze and 

 being uninsulated, this temperature change is probably due 

 partly to conduction when in the warmer water, and partly to 

 the friction of the surrounding water when the bucket is drawn 

 up. The temperature recorded was usually, though probably 

 not always, that of the coldest portion. This temperature, how- 

 ever, may be too large, and there may be some irregularities due 

 to the fact that the coldest portion was not always tested. This 

 difficulty, however, does not affect the accuracy of the density 

 determinations. At the end of each cruise all the samples were 

 taken ashore, and the density of each was tested either by means 

 of a specific gravity bottle, capacity fifty or one hundred cubic 

 centimeters, and a chemical balance or a hydrometer. The 

 specific gravity bottle was of the Gay-Lussac pattern, provided 

 with a ground "lass stopper, having a fine perforation, so that in 

 closing the bottle an excess of water might be forced through 

 the perforation and wiped off with a piece of filter paper. In 

 this way the bottle could be filled to a definite volume, the exact 



volume depending upon the temperatui I' observation. The 



weight and volume being determined, the absolute density was 

 calculated, in ('. G. S. units, the probable error being ± .00002. 

 A set of five hydrometers (by Dr. R. Kuchler, Ilmenau in 

 Thiiringen) was provided, having a range from 1.000 to 1.031, 

 the range of each being .007. The stem was graduated to .0001, 

 and its diameter was about 3.5 mm., the volume of the body 

 being 115 e.c. A series of water samples w r as tested both by 

 the hydrometers and by the specific gravity bottle, the temper- 

 ature of observation varying from 14° to 25 . and a reduction 

 formula deduced for each hydrometer used. By means of a 

 hydrometer the density could be determined with a probable 

 error of ± .00005. 



