292 



REPORT 1851. 



water, as was also the case with experiments Nos. 3 and 4, made on the 31 St 

 of October, when the hour of high water was 10** 30" a.m. 



The sand in both these cases was thoroughly ivel along the base line, but 

 its interstices were not filled with sea-water higher than within perhaps 8 feet 

 of the surface. 



The experiments Nos. 5, 6, 7 and 8, the 2nd of November, when the time 

 of flood tidewas 1 1*^ 57™ A.M.,were,on the contrary,made nearly at high water, 

 and in these cases the sand was saturated to within 2 or 3 feet of the surface. 



Of the above eight experiments the only two upon which the least suspi- 

 cion of inaccuracy can rest, are those of No. 3 and No. 5, in which I am rather 

 doubtful of having stopped my chronograph with absolutely the same promp- 

 titude as in all the rest. 



The error in excess of time, if any, however, could not amount to 3 divisions 

 of the small dial=3 X0"-013914 in time, an amount so small and so much 

 within the probable errors of observation, that I retain all the experiments 

 as good results. 



Deducting column the third in the above table from column the second, we 

 get the following table No. 2, each column being previously brought separately 

 into time, recollecting that the value of one division is different in each chro- 

 nograph. 



Table No. 2. 

 Showing the results of the experiments reduced to time, &c. 



And subtracting now the average of column 4 from that of column 3, we get 

 for the whole time of wave transit — 



3-731303 

 0-819664 



2-91 1639= whole time of wave transit for 



half a mile without correction. 



