Deep temperatures. — For depths to about 30 or 40 fathoms, deep-sea 

 registering thermometers were used. They were carefully checked by comparison 

 with standard thermometers having Kew certificates; and any index error was 

 allowed for. The results to the nearest quarter degree Fahrenheit are thus 

 reliable. This type of thermometer is convenient and satisfactory, so long as 

 the temperature of the water decreases continuously to the depth taken. 



For greater depths, from 50 to 200 fathoms, a deep-sea inverting thermo- 

 meter was used,. This was essential for accuracy where there were layers of 

 water at different temperatures, which was often the case. The thermometer 

 is released by a fan which operates when the instrument begins to be raised 

 through the water. To avoid premature release from the rolling of the vessel, 

 the supporting line was passed over a six-inch pulley wheel, and motion allowed 

 for by hand to keep the thermometer steady in the water at the depth taken, 

 for the necessary time. Any discordant or suspected readings were rejected or 

 the observation was repeated. At 50 fathoms, where the lowest temperature 

 usually occurred, readings were often taken with both types of thermometer 

 as a check. 



Protection of thermometer bulbs. — Both types of thermometers have the 

 usual protecting bulbs to obviate error from pressure at such depths. It had 

 been customary for the outer bulb to be partially filled with alcohol; but in these 

 thermometers, this was replaced by mercury which has a great advantage in 

 transmitting the temperature more quickly to the inner bulb. Instead of keep- 

 ing the thermometer for seven to ten minutes at each depth taken, the time was 

 thus reduced to four or five minutes, which enabled much more work to be done 

 in a day. Also, as the release was by a fan, two thermometers could be placed 

 on the same line, 50 fathoms apart; which enabled certain depths to be checked 

 twice by two different thermometers. Another combination suitable for the 

 lesser depths, was to place a registering and an inverting thermometer on the 

 same line, 10 fathoms apart. 



Densities. — ^The samples were obtained at the depth desired, with a deep- 

 sea water bottle with valves. As the water was often very cold, it was found 

 best to bottle the samples at the time, and to allow their temperature to come 

 to about 60° Fahrenheit before reading the density with a hydrometer. This 

 reduces the temperature correction to a minimum, which is a distinct advantage; 

 because it is large relatively to the variations in density that are being dealt 

 with, and also because the amount of the correction differs appreciably for the 

 various densities from pure sea water to diluted water, as found in the various 

 regions. 



The hydrometers were graduated to show specific gravity; and to obtain as 

 open a scale as possible they were made in series with a small range on each. 

 The total range from fresh water to standard sea water (1 • 0000 to 1 • 0260) was 

 thus substantially subdivided, and adapted to the range of density in the region 

 under investigation in each season. 



Proceedure. — At the Stations where anchorages were made, the current 

 might often be too strong to obtain temperatures and densities vertically down- 

 ward. The tidal fluctuation was therefore taken advantage of, to obtain them 

 when the current was least. In tidal streams which run both ways, they were 

 taken at slack water; as this had the further advantage of showing whether 

 there was any change in temperature with the direction of flow. On the longer 

 temperature sections, there was greater facility, even with current, in measuring 

 depths vertically because the vessel was free. On the other hand, in taking surface 

 temperatures and surface samples for density, it was not necessary to stop. 



Positions. — The location of Stations for anchorage was carefully selected for 

 the purpose in view, and the position was fixed by bearings and angles so that 

 the same spot could be occupied as often as desired. When temperatures and 



6 



