426 DR. F. HORTON ON THE MODULUS OK TOi;s|oN\l, HKHDITV OF 



limits over the part of the tube surrounding the fibre. It was found to be impractic- 

 able to standardise the outer junction by comparison with the inner one, because the 

 difference in their temperatures, although small, varied from time to time, the 

 variation being probably caused by the cooling effects of air currents. 



At the end of the experiments with the quartz fibres the thermocouple was 

 standardised. For this purpose use was made of the determinations of the melting- 

 points of potassium and sodium sulphates, and of the boiling-point of sulphur given by 

 HEYCOCK and NEVILLE.* The platinum tube was supported horizontally, and some 

 very small grains of potassium sulphate were placed upon it in the neighbourhood of 

 the thermocouple. These were viewed by means of a microscope, and the tempera- 

 ture of the tube was gradually increased. The galvanometer deflection corresponding 

 to the melting-point of the salt (1066 C.) was thus determined with accuracy, 

 repeated observations giving identical results. The temperature of the other 

 junctions (of the thermocouple wires with the galvanometer leads) was constant 

 to within '4 C. throughout the whole series of observations, and consequently no 

 correction to the reading had to be made on this account. 



Similar experiments were made with sodium sulphate, and the deflection corre- 

 . spending to 883 C. was obtained. 



In obtaining the deflection corresponding to the boiling-point of sulphur, the 

 junction was removed from the platinum tube and placed inside a thin glass tube, 

 which was then inserted into a retort from which sulphur was distilled. The junction 

 was placed above the boiling liquid and well within the vapour. The barometer was 

 read, and the boiling-point of sulphur calculated from the formula given by HEYCOCK 

 and NEVILLE. The galvanometer deflection corresponding to the boiling-point of 

 water was also determined in a manner similar to that employed with sulphur. 



The galvanometer deflections obtained from these experiments were then plotted 

 against the corresponding temperatures, and from the resulting curve the tempera- 

 ture corresponding to any deflection could be read. 



Three fibres were experimented on over this range of temperature. They were 

 each about 10 centims. in length, and about '006 centim. in diameter. These were 

 made by drawing down the centre of a quartz rod fairly thin, and then heating the 

 thin part and drawing it out suddenly. Practice was required before fibres of the 

 right length and sufficiently thin could be made by this method, but after a few 

 attempts no difficulty was experienced. 



Observations of the period of torsional vibration and of the logarithmic decrement 

 of the amplitudes were taken at intervals of about 50 C. as the temperature was 

 increased from -that of the room to 1000 C., at which temperature the internal 

 friction of the fibres was so great that the vibrations were nearly dead-beat. The 

 fibres experimented on were made from different specimens of quartz, but the results 

 obtained were in every case the same. The rigidity of the fibre increased with the 



* 'Jour. Chem. Soc.,' vol. 67, p. 160. 



