182 WORK OF P. B. DAVIS. 



lation below 35. A closely fitting cover was provided for the conductivity bath 

 for work above 35, to prevent steaming and to maintain the air immediately above 

 the cells at as near as possible the temperature of the water in the bath. 



The temperature of the thermostats was maintained constant to within 0.02 by 

 means of electrically controlled gas-regulators devised by Reid. 1 The thermometer 

 used in the conductivity bath was of the usual 100 inclosed-scale type, graduated in 

 0.1, and could be read to 0.02 with a hand lens. For work up to 45 the viscosity 

 bath was provided with a 25 Beckmann thermometer graduated to 0.05 and with 

 a certified Bender and Hobein thermometer for higher temperatures. All ther- 

 mometers were first standardized against a certified Reichsanstalt instrument, and 

 comparisons were frequently made during the course of the work. 



The conductivity apparatus consisted of the usual Kohlrausch slide-wire bridge, 

 resistance box, induction coil, and telephone receiver. This apparatus was made and 

 standardized by Leeds and Northrup of Philadelphia, and in addition the resistances 

 were compared with a rheostat which had been standardized by the Bureau of Stand- 

 ards, of Washington, D. C. The bridge wire was calibrated as directed by Jones 2 and 

 found to be practically of uniform resistance throughout. Under favorable condi- 

 tions separate readings with the same resistance agreed with one another to 0.5 mm., 

 the nature of the solvent precluding closer agreements. 



The conductivity cells used were of two types: Those of the ordinary plate type 

 described by Jones and Bingham 3 had constants ranging from 40 to 339, and were 

 used with the more concentrated solutions. For the N/10 to N/1600 solutions, 

 five cells of the type previously described by Jones and Schmidt 4 and by Jones and 

 Kreider 5 were used. Their constants ranged from 2.35 to 4.36, and because of this 

 they were well adapted to the measurement of high resistances. 



Cell constants were determined with a fiftieth-normal potassium chloride solution 

 as a basis, this being diluted to N/500 and N/2000 for the low constant cells. 

 Checks were made at regular intervals, and showed only slight variations in the cell 

 constants, especially in the case of the cells with concentric cylindrical electrodes. 



The molecular conductivity of the fiftieth-normal solution was taken as 129.7 

 reciprocal Siemens units, at 25. That of the more dilute solutions was determined 

 by direct measurements. 



Viscosity measurements were made with the Ostwald viscosimeter 6 as modified by 



Jones and Veazey, 7 the sizes of the capillaries being adapted to the solutions for 



which they were employed. The viscosimeters were calibrated as described by 



Jones and Schmidt, 8 the time of flow for water in the instruments used for glycerol 



being derived from the formula 



/ // in 

 tit _ tw ' t A ' Ib 

 l W 77 // 

 I A 'h 



where t' w is the time of flow of water in the water viscosimeter, t' A the time for a slightly 



more viscous solution in the water viscosimeter, t% i' B ' and t' B " the times of flow of 



the solutions A and B of intermediate viscosities between water and glycerol in an 



intermediate viscosimeter and in the instrument for glycerol; tw the derived time 



Amer. Chem. Journ., 41, 148 (1909). Hbid., 4S, 2S2 (1911). 



"Freezing Point, Boiling Point and Conductivity Methods. 6 0swald-Lutker: Physiko-chemische Mess., 3d ed. (1910), 



Amer. Chem. Journ., 34, 481 (1905). p. 232. 



Hbid., 42, 39 (1909). 'Zeit. phys. Chem., 61, 641 (1908). 'Loc. cit. 



