NARRATIVE OF THE CRUISE. 25 



hill on the sides of declivities. The position of the Azores will probahly be found very suitable for observations 

 of this kind. Temperature soundings should be taken at various depths, especially on their north and south 

 slopes, and in the channels between the islands ; and the temperatures at various depths should be compared 

 with those of corresponding depths in the open ocean. 



It is in the Southern Oceans that the study of ocean temperatures at different depths is expected to afford 

 the most important results ; and it should there be systematically prosecuted The great Ice-barrier should be 

 approached as nearly as may be deemed suitable, in a meridian nearly corresponding to the centre of one of the 

 three great Southern Oceans, — say to the south of Kerguelen's Land ; and a line of soundings should be carried 

 north and south as nearly as may be. 



In connection with the limitation of the area and depth of the reef-building corals, it will be very 

 important to ascertain the rate of reduction of temperature from the surface downwards in the region of their 

 greatest activity ; as it has been suggested that the limitation of living reef-builders to 20 fathoms may 

 be a thermal one. 



Wherever any anomaly of temperature presents itself, the condition of such anomaly should, if possible, be 

 ascertained. Thus there is reason to believe that the cause of the temperature of the surface water being below 

 that of the subsurface stratum, in the neighbourhood of melting ice, is that the water cooled by the ice, by 

 admixture with the water derived from its liquefaction, is also rendered less salt, and therefore floats upon the 

 warmer and Salter water beneath. Here the determination of specific gravities will afford the clue. In other 

 instances a warm current may be found beneath a colder stratum; and the use of the " current drag " might show 

 its direction and rate. In other cases, again, it may happen that a warm submarine spring is discharging 

 itself, — as is known to occur near the island of Ascension. In such a case it would be desirable to trace it as 

 nearly as may be to its source, and to ascertain its composition. 



Movements of the Ocean. — The determination of Surface Currents will, of course, be a part of the regular 

 routine, but it is particularly desirable that accurate observations should be made along the line of sounding in 

 the Southern Ocean, as to the existence of what has been described as a general " Southerly set " of oceanic 

 water, the rate of which is probably very slow. It is also very important that endeavours should be made to 

 test by the " current drag," whether any underflow can be shown to exist from either Polar basin towards 

 the Equatorial region. A suitable locality for such experiments in the North Atlantic would probably be the 

 neighbourhood of the Azores, which are in the line of the glacial flow from the North Polar Channel The 

 guide to the depth at which the current drag should be suspended will be furnished by the thermometer, 

 especially where there is any abrupt transition between one stratum and another. It would be desirable that 

 not only the rate and direction of surface drift, but those of the subsurface stratum at (say) 200 fathoms' depth, 

 should be determined at the same time with those of the deep stratum. 



Tidal Observations. — No opportunity of making tidal observations should be lost. Careful observations 

 made by aid of a properly placed tide-pole in any part of the world will be valuable. Accurate measure- 

 ments of the sea level once every hour (best every lunar hour, i.e., at intervals of l h 2 m of solar time) for a lunar 

 fortnight (the time of course being kept) would be very valuable information. 



Bench Marks. — In reference to the interesting question of the elevation or subsidence of land, it will be 

 very desirable, when sufficient tidal observations can be obtained to settle the mean level of the sea, that 

 permanent bench marks should be established, recording the date and height above such mean level. Even 

 recording the height to which the tide rose on a certain day and time would render a comparison possible in 

 future years. 



A good determination of the mean sea level by the simple operation of taking means may be made, in less 

 than two days, with even a moderate number of observations properly distributed so as to subdivide both solar 

 and lunar days into not less than three equal parts. Suppose, for example, we choose 8-hour intervals, both 

 solar and lunar. Take a lunar day at 24 h 48 m solar time, which is near enough, and is convenient for division, 



(narb. chall. exp. — vol. i. — 1884.) 4 



