Palmer — Pressure Coefficient of Mercury Resistance. 3 



closed at the ends by rubber stoppers, through which water 

 from the city mains was allowed to flow continuously. A ther- 

 mometer with its bulb inside of this cooler showed that the 

 temperature never varied more than one degree from 9° C. 

 throughout the entire series of experiments and that the varia- 

 tions during the same day were very much less than this. For 

 the measurements at the boiling point of water a tin can about 

 30 cm long and 13 cm in diameter, having short brass tubes sol- 

 dered to apertures in the center of its ends, was placed on the 

 piezometer and fastened, by short pieces of rubber tubing, in 

 such a position that it entirely covered the experimental «tube 

 within. Two openings in the top were provided, one for the 

 reception of a thermometer and the other for a vertical water 

 condenser. The latter, being open at the top, kept the steam 

 at atmospheric pressure and at the same time obviated the 

 necessity of frequently renewing the supply of water. The 

 whole arrangement with the exception of the bottom was 

 covered with asbestos to prevent radiation and heat was applied 

 by means of a ring burner surrounding the piezometer below 

 the can. Small water coolers were placed above and below to 

 prevent the conduction of heat through the steel tube to the 

 joints where it might cause leaks. 



Various methods for the measurement of resistance were 

 tried with more or less success, but that due to Carey Foster 

 was found to give the best results owing to its sensitiveness 

 to small variations. The general arrangement of the apparatus 

 for this method is too well known to need description here. 

 The transposition of the standard and unknown resistances 

 was accomplished by means of an eight pole mercury commu- 

 tator similar to those put on the market by Raider Bros. A 

 series of platinoid coils, by Queen & Co., so arranged that their 

 combined resistance could be varied by tenths from zero to 

 ten thousand ohms, without altering the number of plugs in 

 the circuit, was used as a standard with which to compare the 

 mercury thread under investigation. As noted above, the 

 electrodes of the thread were connected respectively to the 

 piezometer and compressor, and since these parts were other- 

 wise very perfectly insulated from one another they served 

 admirably as poles from which to make connection with the 

 commutator. A very uniform german silver wire, about No. 

 17 B. & S. gauge, was wound in ten uniform spirals about a 

 vulcanite cylinder 10 cm in diameter and 3*6 cm long. Its ends 

 were fastened, in the same generating line of the cylinder, to 

 two thick brass plates that formed the ends of the drum and 

 were rigidly fastened to two stout brass pillars which were con- 

 nected with the poles of the commutator. An insulated frame 

 work was arranged to revolve about this drum in such a man- 



