168 On the Thermoelectric Position of Platinoid, 



electric distance of platinoid from Professor Tait's lead wire 

 -1128-4-lx*. 



This result enables ns to place platinoid in Tait's thermo- 

 electric diagram. Its line is nearly parallel to those of palla- 

 dium and German silver, and slightly above the latter. It is, 

 however, to be remembered that, in all probability, different 

 specimens of platinoid alloy would give results differing con- 

 siderably from that quoted above. 



Appendix. By A. Tanakadate. 



The following experiment on the torsional rigidity of spider- 

 line was carried out in the Physical Laboratory of the Imperial 

 University of Japan in 1884, and a notice of it was published 

 in vol. ii. of Rigakukyokwai Zassi (' Proceedings of the Science 

 Society') of that year in Japanese. It has not hitherto been 

 described in English ; and the absolute determination, as 

 referred to below by Mr. T. Gray, of the rigidity of silk fibre 

 makes an estimate of the rigidity of spider-line possible. 



The determination of the torsional rigidity was a relat : ve 

 one, and the experiment essentially consisted in finding the 

 deflexion of a small magnet due to a given twist of the sus- 

 pending fibre ; the magnet being placed in the earth's mag- 

 netic field (0"3 C.G.S.). The deflexion was observed by the 

 usual method of the reflected image of a fine wire stretched 

 before a lamp. 



The mirror-magnet was first hung by a silk fibre of 31 cen- 

 tim. length and placed in the usual way. The distance of 

 the scale from the mirror was 95 centim. When the torsion- 

 head of the magnetometer was turned through one complete 

 revolution (27r) in either direction from zero, the image of 

 the reflected wire was displaced through 8 millim. either wav 

 or -8/2 x 95 =0-0042 radians, or 864". "' ' 



The silk fibre was now detached from the magnet and a 

 spider's line (newly spun) was attached in its stead. The 

 length was 28 centim., the magnetometer was put into its 

 place, and the torsion-head was turned as before ; but no ap- 

 preciable deflexion could be observed, even when the torsion- 

 head was turned through ten complete turns (20-7r) . It was 

 then suspected that the mirror might have been caught against 

 the sides of its case ; a close inspection, however, showed that 

 it was quite free. The fibre was then shortened to 2*3 centim. 

 (about one twelfth its previous length) and the experiment 

 was repeated. Ten complete turns of the torsion- head gave a 

 deflexion of 1*5 millim. ; or • 1 5/2 x 95 = 0*00079 radians = 16"*3 

 per turn. 



