90 Mr. T. G. Bedford on the Expansion of 



But there are one or two experiments that might be suggested 

 that could be very easily performed without any special 

 expensive or complicated apparatus. Thus it would be very 

 interesting to keep an iron plate red hot in a stream of pure 

 dry hydrogen for several hours, allow it to cool in that gas, 

 leaving it say for a few days, and then measure the contact P.D. 

 between it and another iron plate that had not been so 

 treated. This might be repeated with a platinum plate, and 

 the two experiments together would then be somewhat equi- 

 valent to those described on pp. 85-87, and very much 

 simpler to perform. 



§ 17. It is with the hope that somebody will find it worth while 

 to continue investigations along these lines, that I venture to 

 publish these preliminary results *. 



Throughout the whole of this work I have been fortunate 

 enough to receive the valued assistance and advice of 

 Prof. W. E. Ayrton, F.RS., in whose laboratories at the 

 Central Technical College the work was done, and of 

 Mr. T. Mather, his Senior Demonstrator in the Physical 

 Department. I must also thank my friend Mr. W. Duddell 

 and Mr. W. T. Evans, who have so often kindly helped me, 

 when necessary, in the experimental part of the work. 



V. On the Expansion of Porcelain ivith Rise of Temperature. 

 By T. G. Bedford, B. A., Sidney Sussex College, Cambridge^. 



IN experiments involving the use of air-thermometers, 

 such as comparisons between the air and platinum- 

 resistance scales of temperature, a correction is required for 

 the expansion of the thermometric envelope. Over a range 

 at low temperatures the expansion might be readily deter- 

 mined, e. g. y by filling the bulb with a liquid, conveniently 

 mercury, whose absolute expansion is known, and using it as 

 a weight-thermometer to measure the relative expansion of 

 the liquid. At high temperatures, however, such a method 

 cannot well be used, and the correction must be calculated 

 either from the results of low temperature measurements, or 

 from the linear expansion as determined directly at these 

 higher temperatures, provided the material be homogeneous. 

 As porcelain is the substance generally used, this latter 

 method is available; and it was therefore suggested to me by 

 Mr. E. H. Griffiths, F.R.S., that it might be useful to try 



* Since the above was written I am very pleased to be able to add that 

 Mr. H. Tomlinson has now taken up the thread of this research, working, 

 for the present, with my old apparatus, and starting from the point at 

 which I left off. 



t Communicated by the Physical Society. 



