HYDROGRAPHY. 



95 



requires to make it fully saturated, must be regarded as a measure for the estima- 

 tion of the reducing effect the water has been subject to, since the last time it left 

 the surface. 



As it appeared to me, that these theorems had been used without any clear understanding of 

 the correctness of the results gained by them, I have, as will be seen from the table, made a great 

 deal of experiments to determine the quantity of gases held in surface-water. 



In case of the two theorems being fully correct, they will require that surface-water should con- 

 tain exactly the quantity of nitrogen and oxygen required to saturate it. This is, as will be seen from 

 the table, not quite correct, and in consequence of this, we may expect to make as great errors by 

 the application of the theorems on water from the depth, as when applied to surface-water. The 

 theorems must therefore be modified, but as it is such simple physical causes, to which, at least in so 

 far as regards the quantity of nitrogen, these modifications are due, we shall by these causes have new 

 means to ascertain the anterior state of the water. 



Several determinations of alkalinity and sulphuric acid, as well as a few determina- 

 tions of lime and magnesia, have been made with the water-samples brought home from the 

 voyage in 1895. The alkalinity determinations for 1895 have — like similar determinations for 1896 — 

 been made with water-samples deposited in glass-bottles, so that the exactness of these determinations 

 cannot be much relied upon; most unfortunately it came to my knowledge too late that sea-water, 

 when subject to storing, had such an influence on glass, that it could dissolve perceptible quantities 

 of the alkali of this latter, which makes the alkalinity determinations unreliable. Notwithstanding this 

 fact, the relative values may still be worth taking into consideration, and in the following table the 

 alkalinity is put down expressed by cubic centimetres of carbonic acid per litre. 



