148 HYDROGRAPHY. 



height By the temperature of absorption is understood the temperature the water is required to 

 have at a pressure of 76o ram to be saturated with the amount of nitrogen it is found to contain. 



To answer the question before us, we must have ascertained by which of the two quantities 

 (c — a) and (c — b) the most homogeneous results can be attained. Looking at the table, it will be seen 

 that (c — a) shows, on an average, a more reliable dependence on the heating than c — b. Station no 

 is apparently an exception to this, as for a heating of 6°8 we have c — a = o-44 ccm , c — b = o-84 ccm , 

 while station 101 for a heating of 6°o gives c — a = 0-93 ccm , c — b — o.95 ccm . It can be seen here at 

 once, however, that the measure used for heating proves to be inappropriate, as the two stations have 

 nearly the same surface-temperature, notwithstanding the observations at station no being taken 10 

 days later than those at station 101. Station 121 where the observations again have been taken 19 

 days later than those at station no, has for a heating of 5°i c — a =-- 0-30, c — b = o - 2i. Consequently 

 station no does not refute the aforesaid assumption, but, on the other hand, it does not prove any- 

 thing at all, for the state of matters found at the stations 101, no and 121, could very well be due 

 to a continued minimum of the barometer-height at these places between the moments of time, at 

 which the observations at the stations 101, no and 121 were taken. The other stations where the 

 observations have been taken under a conspicuous high or low height of barometer, are lying in too 

 isolated positions to allow us to draw any reliable conclusions concerning the aforesaid assumption, 

 but, as has been said before, c—a seems, upon the whole, to be regulated by the heating with greater 

 precision than c — b. 



The consequence of this is that we must be justified in not having regard to the 

 pressure of air during the absorption, when by means of the gases contained in the 

 water we are going to determine the temperature at which the absorption has taken 

 place. Such a determination is, as it will be seen, anything but reliable, as surface-water may con- 

 tain up to i ccm of nitrogen more per litre than it ought to do according to the law of absorption, so 

 that on account of this, the calculation may be subject to an error of 3 . Hence the temperatures 

 found in this manner, do not with any high degree of exactness represent the temperature of the 

 water when it left the surface. 



The following two circumstances may be the cause of water ceasing to be surface-water. It 

 will either on account of cooling obtain a higher specific gravity than the stratum of water below it, 

 and in consequence of this have a motion towards the depth, or, as another possibility, it is not 

 excluded that a layer of less specific gravity could be flowing and spreading itself on top of it 



In the first case, it may be expected that the water is not saturated with nitrogen when it 

 leaves the surface. The fact is that there is every reason to believe, that as the water is supersatu- 

 rated during the heating, it will not, when subject to cooling, be saturated with the said gas. If 

 now such water is not subject to heating on its way to the depth, it will for all that constantly re- 

 main in an unsaturated state, and we cannot therefore from an observed slight incomplete 

 saturation draw the conclusion that the water has been subject to cooling, while, 

 on the contrary, a supersaturation proves with certainty that the water has been 

 subject to heating after it left the surface. 



