2 Scientific PrucceiNngs. Roi/dl Dnhlin Socipii). 



Apparatus. 



'i'he construction of the tliermo-couples employed, and the general 

 arrangements of the thermo-electric method, have already heen described (2i, 

 and it is unnecessary to go into them here. Slight modifications have, 

 however, been introduced. Instead of single drawn pine rods to support the 

 wires of the couple, double ones, slightly flattened on the surfaces of contact, 

 were employed. This gives greater strength, and affords a means of pro- 

 tecting the eureka and copper wires forming the couple, which run down the 

 grooves on each side of the double pine supports. One of the pieces of pine 

 is cut away close to the lower end of the support, thus forming a recess in 

 which the junction is freely exposed to the surrounding fluid. Collodion 

 varnish was used Lo waterproof the insulated wires, and to cement the pine 

 rods together, and to keep the wires in position. In case it is desired to 

 alter the sensitiveness of the couple, or modify the couple in any way, it is 

 easy to dissolve off the collodion with acetone, and substitute a new eureka 

 wire of different resistance. 



Calibration of the Thermo-coitples. 



Sucrose was employed as a standard for the calibration of the thermo- 

 couples. It was selected for this purpose, not only because it can be readily 

 obtained of great purity, but its depressions of freezing-point, which are 

 known with great precision from Eaoult's researches, can be more easily 

 determined than those of other substances, whose solutions are not hydrated 

 to the same degree. The ease with which it undergoes inversion on storage 

 is, however, a serious disadvantage to its use ; and as it is not always pos- 

 sible to make up standard solutions for every calibration, calcium chloride 

 solutions of suitable concentration, whose freezing-points had been deter- 

 mined against sucrose, were used for this purpose. 



The importance of frequently calibrating the instrument has already 

 been emphasized (2). 



Thermo-couples producing deflections of from two to three hundred milli- 

 metres per 1° C. have been constriicted with a single pair of junctions, the 

 degree of sensitiveness varying with the resistance of the eureka wire 

 employed. The calibration of a thermo-couple producing a deflection of 

 294'5 mm. per 1° C. is shown in Table I. In this approximately oO cm. of 

 eureka wire, 0100 mm. diam., and two leads each of copper wire, 0-200 mm. 

 diam., were used. In the last column of the Table are given the actual 

 freezing-points of the solutions derived from Eaoult's results Cjuoted by 

 Hamburger (4). 



