546 REPORT — 1901. 



to be worked at the lamp, and then to secure that the selected area shall occupy 

 the desired position in the completed apparatus. 



To secure precision many precautions were taken. The construction and 

 mountino- of the instrument is much like that of an astronomical instrument. A 

 massive cast-iron standard, designed so as not to he distorted by changes of tem- 

 perature, rests on an isolated stone pier; on it, moving on trunnions in V-shaped 

 supports like those of a transit instrument, is carried the plate on which siphon 

 gauge, mirrors, and microscope are fixed. The free surface of the mercury in each 

 arm of the gauge is five centimetres in diameter. The tube connecting them is two 

 centimetres in diameter, and is but two centimetres below the free surface ; so that 

 the temperature of the two columns of mercury shall be equalised rapidly. Good 

 illumination is provided, with care to minimise the access of heat to the mercury. 

 The pair of mirrors is provided with every motion required to bring the two fidu- 

 cial points into focus at once, and to give the images of the points any desirable 

 position in the field of the microscope. The ends of the points are wrought into 

 small hemispheres. With all these precautions, as well as many others, we have 

 been able to make measurements in which the mean error of a single reading is not 

 very much greater than a ten-thousandth of a millimetre. 



4. The Transmission of Heat through Water Vapour. 

 By Charles F. Brush and Professor Edward W. Morley. 



In the discussion which was elicited by the paper of Mr. Brush on a new gas 

 whose power of transmitting heat is vastly greater than that of hydrogen, Sir 

 "William Crookes suggested that the observed phenomena might perhaps be due to 

 water vapour, and described experiments which seemed to ' show that, at high 

 vacua, water-gas is a better conductor of heat than either air or hydrogen at similar 

 pressures.' 



Being able now to measure small pressures directly, we have determined the 

 rate of transmission of heat through water vapour at pressures from that of satura- 

 tion at 0° to less than a millionth of an atmosphere. The three gauges described 

 before have been used in three series of experiments with three different apparatus. 



At low pressures, water vapour transmits heat more rapidly than air, but not 

 so rapidly as hydrogen. The superiority over air is a maximum at twenty or thirty 

 millionths of an atmosphere, and is not far from 30 per cent. At sixty or 

 eighty millionths, air and water vapour transmit heat at the same rate ; at higher 

 pressures, water vapour transmits heat less rapidly than air at the same pressures. 

 Statements more precise than these cannot now be made, because the form and 

 dimensions of the apparatus used modify slightly the curves which represent the 

 relations between pressure and rate of transmitting heat, and the place of intersec- 

 tion of the curves is therefore uncertain. 



5, Comparison oj the Constant Volume and Constant Pressure Scales for 

 Hydrogen between 0° C. and — 190° C. By Morris W. Travers, D.Sc, 

 and George Senter, B.Sc. 



The authors describe a modified form of constant volume gas thermometer in 

 which the average temperature of the stem — the part connecting the bulb with 

 the so-called ' dead space ' — is determined from the readings of a secondary gas 

 thermometer the bulb of which lies side by side with the stem of the main 

 thermometer. The relation between the two scales was deduced from the expan- 

 sion of hydrogen at constant pressure between —190° C. and 0° C. The arrange- 

 ments used to determine this expansion were as follows: — The bulb of the 

 constant volume thermometer was immersed in liquid air side by side with 

 another bulb, which we may call the constant pressure bulb, filled with hydrogen 

 at a known pressure, the temperature being deduced from the readings of the 

 constant volume thermometer. The gas in the constant pressure bulb was then 



