688 REPORT — 1897. 



tiyities of the constituents, taken in the proportions in which they occur in air 

 differed from that of air by being 0-35 per cent, too small. Similar experiments 

 made -with a mixture of hydrogen and helium gave the result that the sum of the 

 refractivities of these gases, taken separately, differed from that of the mixture by 

 no less than 3 per cent, in excess. It is thus probable that gases are not without 

 influence on each other, but that in some cases the refractivity is diminished, in 

 others increased by mixture. 



2. Note on the Use of the Trifilar Suspension in Physical Apparatus. 

 By SiLVANUs P. Thompson, F.B.S. 



The author advocated the use of trifilar suspensions in certain forms of 

 apparatus, as having the advantage over bifilar in not being liable to be thrown 

 into lateral pendular motion. He instanced the case of a differential D'Arsonval 

 galvanometer, of a moment-of-inertia apparatus designed by Professor Daiby, and 

 of an apparatus designed by himself for illustrating mechanically the transmission 

 of transverse vibrations and of Hertz waves, in which model the part corresponding 

 to the Hertz resonator (a metal ring) was hung by a trifilar suspension. 



3. On Zeeman's Z>iscovery of the Effects of Magnetism on Spectral Lines. 

 By Professor O. J. Lodge, F.B.S. 



4. On the Use of a Constant Total Current Shunt zvith Ballistic Galvano- 



meters. By Professor W. E. Ayrton, F.B.S., and J. Mather. 



5. The Sensibility of Galvanometers. By Professor AV. E. Ayrton, F.B.S., 



and J. Mather. 



6. Short versus Long Galvanometers for Very Sensitive Zero Tests. 

 By Professor W. E. Ayrton, F.B.S., and J. Mather. 



7. On a Besearch in Thermo-electricity hy means of a Platinum Besistance 

 Pyrometer. By H. M. Tory, M.A., Lecturer in Mathematics and 

 Demonstrator in Physics, McGill University, Montreal. 



[Communicated by Prof. H. L. Callendah, M.A., F.R.S.] 



The paper is an account of some experiments carried on in the McDonald 

 Physics Building of McGill College, with a view to applying the electrical resist- 

 ance pyrometer to the phenomena of thermo-electricity. 



The method was suggested by Professor Callendar, whose work, with that of 

 Messrs. Griffiths, Heycock, and Neville, has fairly well established the formulas 

 for calculating temperature by this means. 



The object of the investigation was to give a more rigid verification of Tait'a 

 parabolic formula. 



The research, as conducted, naturally divides itself into three parts :— 



1. The study of the usual form of copper-iron junction. 



2. The study of a cast-iron wrought-iron junction. In conducting some experi- 

 ments on the cyclical variation in the cylinder wall of a steam-engine. Professor 

 Callendar found a couple of this type most suitable. 



3. A direct comparison of the electrical resistance pyrometer with the platinum- 

 platinum-rhodium couple. 



In all cases the compensation method was used for measuring the E.M.F., 

 the junction being balanced against a storage cell, which in turn was continuously 

 balanced against a Clark cell kept at constant temperature. A carefully calibrated 

 rheostat and a resistance box, both of the same material, were used. 



