52 R. A. GORTNER AND J. A. HARRIS 



numbered by one of us and placed in ice water for preliminary 

 cooling, by having the freezing mixture vessel of the thermometer 

 insulated, by having an auxiliary jar with ice and salt at about 

 -4° for cooling the sample (transferred to the freezing tube) 

 down to near its freezing point^ before placing in the air jacket, 

 and by using a steel air jacket*' instead of the conventional glass 

 tubes which are apt to be broken by rapid work. After cooling 

 to near its freezing point in the auxiliary bath the freezing tube 

 (containing the thermometer bulb) was quickly dried by filter 

 paper and transferred to the steel jacket, the air in which was 

 kept at a low temperature by corking the tube when not in use. 



The freezing point of double distilled water, or better of con- 

 ductivity water, should be redetermined at the beginning and 

 end of each day's work, for in addition to the secular change 

 possibly correctly explained by Atkins^ as due to the distilla- 

 tion of mercury from the more convex surface to the more plane 

 surface beneath in the upper reservoii', there may be a material 

 change over night when the instrument has been constantly 

 used during the day. Probably this is due to ^'ariations in the 

 wall of the bulb which slightly alter its volume.^ 



The methods here described are the outcome of the experience 



ing has no advantages when one is working with vegetable saps, is often in the 

 way, and causes trouble in cleaning and drying the tubes. So we have had tubes 

 an inch shorter than the standard ones, and lacking the useless accessory, made 

 to order (at about half the price of the others) and have found them very con- 

 venient. 



* To avoid too great variations in the wall of the thermometer bulb this cooling 

 (which draws a relatively large volume of mercury from the upper chamber 

 through a fine capillary tube) must not be carried out with too great rapidity — 

 hence the importance of having this bath at only -2° to -4°. Because of the air 

 jacket and because of the fact that the mercury is already low in the capillary, 

 the insulated Beckmann jar may be kept at -15° or lower. 



^ We have found the steel jackets for the 100 cc. centrifuge tubes of the Inter- 

 national Instrument Company to be admirable for this purpose. 



7 Atkins, W. R. G. Sci. Proc. Roy. Dublin Soc. n. s. 12: 124. 1909. 



^ Our maximum difference in the zero point of the instrument as observed at 

 the beginning and end of the day's determinations does not exceed 0.001° or 

 0.002° C. while the change from one evening to the next noon (the instrument 

 standing at laboratory temperature) has in some cases been 0.01° or slightly more. 

 Errors of determinations of the zero point on the instrument of 0.001° to 0.002° 

 might easily be made, but discrepancies as large as 0.01° cannot be thus explained. 



