tlic ])iil)lislu'(I salinillos. Of tlic \';U'ioiis ]);iitial and coinplelc calibra 

 tioii iiicasurcMucnls made in the postwar period prior to February 

 lOOO, those made in FebriiarxMarch 194S, are considered the most 

 reliable and liaxf been accepted. The ciunidative discrepancy of these 

 iiieasurenieiits from the Febriiai'v 19()0, nieasiiremeiits, as obtained 

 from the snccessive interconiparisoii of different batches of Copen- 

 hagen stanchird water, amounts to 0.085 in X-dial I'eading-. Tlie recon- 

 cihation of this dilference between the two ends of the 12-year |)ei'iod 

 may be tieated indifferent ways. It has been assnnied here that this 

 dilference occin'red gra(hially, at the rale of 0.007 pel' year (ca. 

 0.001°/ooS) fi-om 1048 to 19()(),and that thisih'ifl may rei)resent a slow- 

 increase in conductivity of tlie Copenhagen water from solution of tlie 

 glass containing-ampules duiing stoi'age. Three diffei-ent calibration 

 cui'ves wei'e used dnring the ])erio(l ; the 1948 curve foi' 1948-49, the 

 1950 curve for 1950-50 and the 1957 curve for 1957 and sul)sequent 

 measurements. Th(> ai'l)itrai'ily selected point was difl'ercMit for each 

 of the curves so that a salinity' of ;i5°/oo corres|)Oii(h'd to the following 

 X-dial reacbngs: 1948 curve, 50.255; 1950 curve 50.100; 1957 curve, 

 49.911. The signilicant chronology of the interc()mi)ai'ison of the 

 different batches of ( 'opeidiagen walei', in X-dial reading, was as 

 follows . 



Fcbniary-AIarch l(t48 P15 = r)().() 10 



Season of 1051 IM S uivatcr Ihan IM.") by 0.1 tO 



March 1953 PI? f^ivatcr tiian V\n ]>y 0.135 



Season of 1956 P22 greater Ihan 1M7 l)y 0.120 



March-May 1957 P22 greater Ihan P23 by 0.009 



February 19C)0_ P23 = 49.857 



Kcliruary 19()0 _ P29== 49.902 



From t he foregoing, then, the corrected vjdues of X-dial reading which 

 should have been used in st andardiza I ion for the a|)|)ropriale calibra- 

 tion cur\'e used, are as follows: 



19 IS curve 19 18 P15 = 50.0 10 



1 94.S curve 1 949 P15 = 50.033 



1950 curve t950 P15 = 49.871 



19.")0 curve 195t_ P15 = 49.8()4 



1 950 curve 1 95 1 P18= 50.004 



1 950 curvi' 1 952 P15 = 49.857 



1950 curve 1952 P18= 49.997 



1950 curve 1953 P15= 49.850 



1950 curv<' P,)53 P17 = 49.985 



1950 curve 1954 P17 = 49.978 



1 950 curve 1 955 PI 7 = 49.971 



1950 cur\c 195() - P17 = 49.964 



1 950 curve 1 95(i P22= 50.084 



1957 curve 1957 P22 = 49.888 



1957 curve 1957 P23 = 49.879 



1957 curve 1958 1^23 = 4!). 872 



1957 curve 1959 P23 = 49.865 



1957 curve 19()() P23= 49.857 



1957 curve 1960 P29= 49.902 



38 



