SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 6o 



many cases to reject one of them as probably less accurate than the 

 other. Where two readings were taken those which are probably 

 the most nearly correct are (discarding the insignificant fifth deci- 

 mal) as follows : 



No. i 

 No. 2 

 No. 4 

 No. 5 

 No. 7 

 No. 9 

 No. io 

 No. ii 

 No. 12 

 No. 13 

 No. 18 

 No. 19 

 No. 22 

 No. 24 

 No. 25 

 No. 27 

 No. 28 

 No. 29 

 No. 30 

 No. 32 

 No. 40 

 No. 41 

 No. 42 

 No. 43 

 No. 46 



1.0212 

 1.0257 

 1.0259 

 1.0256 

 1.0257 

 1.0251 

 1.0252 

 1.02 18 

 1.0252 

 1.0243 

 1.0244 

 1.0250 

 1.0250 

 1.0250 

 1.0250 

 1.0254 

 1.0253 

 1.0252 

 1.0252 

 1.0247 

 1.0253 

 1.0254 

 1.0254 

 1.0253 

 1.0253 



CALCULATION OF THE CONSTANT REPRESENTING THE 

 INSTRUMENTAL ERROR 



During the cruise of the *' Albatross " among the Philippine Islands 

 two hundred and twenty-one observations of the surface density 

 were taken, with the instruments which I myself had used in 1906; 

 these observations show considerable variation, but the average is 

 1.024533. 



There were also taken on the same cruise eighty observations of 

 the salinity below 100 fathoms ; these likewise exhibit a considerable 

 amount of variation, and give an average of 1.025048. 



The records for the eighty-four observations taken among the 

 Philippine Islands below the surface are in detail as follows: 



