221 Mr. K. Tsnruta on some Effects of 



subject-matters — for example, G. Meyer's in Wiedemann's 

 Annalen, and Gr. Maclean's in the ' Proceedings of the Royal 

 Society of London/ Having been thus reminded of my old 

 experiments, I have been led to do my best to give now and 

 here an account of them, inasmuch as I believe that though 

 unfinished and incomplete, yet if communicated to fellow- 

 workers interested in such matters, they may not be deemed 

 utterly worthless. In undertaking this task at a date remote 

 from the time of actual experiments J cannot feel myself sure 

 of being free from consequences arising out of a long dis- 

 missal of the subject-matter, although my laboratory -journal, 

 to collect my materials from, has been kept unimpaired, and 

 the general features of my result were at that time carefully 

 written out. The publication of this confessedly incomplete 

 account will, I hope, call forth my colleagues' criticism and 

 correction, affording, I doubt not, what 1 myself could not 

 till now, 



G. Wiedemann, in a paper on u Magnetische Untersuch- 

 ungen/' after remarking that as E. (John has found a thermo- 

 electric current produced in a wire, a part of which is 

 stretched and the remainder is not, so the same might take 

 place in regard to twisting, goes on to say that he has con- 

 firmed this by experiment, but it is not to be ascertained how 

 he did so, At the time of my experiments I could find no 

 paper in which the present subject is systematically treated of. 



The following was one of the arrangements used in my 

 experiments of 1892, and may be considered as corres- 

 ponding to the well-known one of Lord Kelvin's in regard 

 to the effect of longitudinal tension. A long piece of wire to 

 be examined, being passed through a brass bar, was clamped 

 to it at its middle part. One half of the wire could be twisted 

 by means of a torsion-head at its free end, the whole being- 

 stretched by means of leaden weights applied to the other 

 free end. A pair of small circular brass disks was fixed on 

 the wire near its free ends and dipped into mercury pools 

 which were connected with the electrodes of a small galvan- 

 ometer. Heat was to be applied at the clamped middle of 

 the wire. This arrangement, simple and good in principle, 

 presented many practical difficulties, and so no quantitative 

 observations were made. 



The arrangement adopted in the year 1895 corresponds in 

 principle to that adopted by E. Cohn in his investigation on 

 the thermo-electric effects of lonoitudinal tension. A torsion- 

 head used was constructed in the following way : — Its outer 

 cylindrical piece, which was to be fixed, had an inner piece 

 (also cylindrical) fitting accurately and easy to turn. This 



