132 



RECORD AND RESULTS OF 



The second part of Table I contains the observed times of high and low water, 

 corrected for error of watch. The adopted watch corrections for June 22d, 23d, and 

 24th, were +18, +15, and +12"' respectively. For June 29th, the correction was 

 + 10"': and for July 11th and 12th, +33 and +U"\ 



Determination of the Mean Level of the Sea. 



An inquiry into the reading of the mean level of the sea is important in more 

 than one aspect ; first, we may test the value of our observations with respect to 

 the invariability of the zero point of the scale which may change from the following 

 causes : a gradual lengthening of the rope ; a gradual shifting of the weight by 

 which it is anchored on the slojnng bottom by the action of currents, or by ice, and 

 possibly also by a motion in the ice-field itself upon which the tidal apparatus rested, 

 and finally by a change in the position of tlie weight after the rope had been taken 

 up for repairs and was replaced. Secondly, by marking, at certain epochs, the half- 

 tide level of the sea, which is subject to smaller fluctuations than either the average 

 level of high water or the average level of low water, we may ascertain any relative 

 change in the level of sea and land produced by geological causes. All levelling 

 operations must also be referred to a certain tidal level. Thirdly, since theory 

 points out certain fluctuations in the tidal level of the ocean due to the diff'erential 

 action of the sun and moon, their study and comparison with observation will bring 

 them to a practical test. There are other interesting questions connected with the 

 subject of our inquiry, namely, the eff'ect upon the level of the sea, of a change in 

 the atmospheric pressui-e as indicated by the readings of the barometer, and also 

 the efi'ect of the wind, with respect to direction, duration, and strength, upon the 

 average height of the sea. The change of the sea level for a given rise or fall of the 

 barometer has only been ascertained for a few places, and the measures fail to show 

 a satisfactory agreement. The effect of the wind is of an entirely local character. 



The mean, or more properly the half-tide level, is the one to which all heights 

 should be referred ; on the average, therefore, we will have at high tide an equal 

 sectional area of water above, and at low tide an equal sectional area of deficiency. 

 Owing to the daily inequality and the half-monthly inequality, which have to be 

 eliminated, the following process for finding the half-tide level was employed. 



