L. A. Bauer — Ocean Gravity Observations. 247 



were ascertained only once. Thus the comparison of Hecker's 

 thermometers with the standard of the Physikalische Reichs- 

 anstalt, or with any other standard, were never again repeated 

 as far as known. Hecker does not give the actual observed 

 corrections, but, instead, a table computed therefrom, which, 

 except for very slight corrections due to the second determi- 

 nation of the calibration corrections, is used for the three 

 cruises. Hecker assumed that the variations in the mentioned 

 corrections, with age and use of thermometers, would either be 

 negligible or "give only a constant difference," and hence 

 enter into the miscellaneous constant term of his observation 

 equation. How justifiable these assumptions are I leave others 

 to judge. Due to the severe conditions imposed by his stren- 

 uous program, Hecker had serious trouble at times with his 

 thermometers — sufficient, indeed, to require rejection of some 

 series. The caliber corrections, I am informed by thermometry 

 experts, are the ones least liable to appreciable changes, and 

 from their re-determination no certain conclusions can be 

 drawn as to the behavior of the other and more important 

 corrections. 



Since Hecker criticises our proposal as to the necessity of 

 frequent controls of the zero point, it will be of interest to 

 quote from such an eminent authority on precise thermometry 

 as Professor Callendar : " The effect [of zero changes] cannot 

 be calculated or predicted in any series of observations because 

 it depends in so complicated a manner on the past history and 

 on the time. It is a most serious difficulty in accurate mercu- 

 rial thermometry, especially at high temperatures. The most 

 satisfactory method of correction appears to be to observe the 

 zero immediately after each reading and to reckon the temper- 

 ature from the variable zero thus observed." The various 

 experts consulted in this country are in entire accord with Pro- 

 fessor Callendar. Now this is what Hecker says (I. o. 9 p. 392) : 

 "The reason why I made no freezing-point observations is that 

 they would have introduced new errors into the observations ; 

 for freezing-point observations are also subject to errors." 

 Experienced physicists would say that the neglect of the zero 

 control introduced greater uncertainty than that of a zero 

 determination. Hecker depended too much upon the possibil- 

 ity of eliminating all outstanding evils by general least square 

 adjustments ; this same remark applies to other matters referred 

 to in his comments, e. g., barometer corrections. 



A word with regard to Hecker's least square treatment of his 

 observational quantities. While I have pointed out wherein 

 his observational work was in some respects not wholly satis- 

 factory, I am inclined to think that the error due to reduction 

 will be found to be greater than the purely observational one. 



