396 Messrs. K. Honda and S. Shimizu on the Change of 



and 1*5 cm. in soft iron ; hence it can be easily judged how 

 great difficulty is experienced in obtaining a constant tempe- 

 rature. Since the demagnetization was carefully effected 

 before each observation and then an instantaneous deflexion 

 was noted, a slow displacement of the image, such as 1 mm. 

 in several minutes, could not cause any sensible error in the 

 results. In our experiments, during the whole set of obser- 

 vations which usually required ten or fifteen minutes, the 

 displacement of the zero point was only two or three milli- 

 metres, except in a few cases. 



After one set of obseivations had been taken, a current of 

 increased strength was again supplied by the dynamo for about 

 an hour or more till the temperature became nearly constant. 

 The current-supply from the dynamo was then changed for 

 that from the accumulator to repeat the same subsequent 

 processes. In this way experiments at successively increasing- 

 tempera ttires from the ordinary to the highest were performed. 

 Our heating-coil gave a rise of temperature of about 1000^ C. 

 in the core by passing a current of 11 amperes. 



When the specimen was once heated to a high temperature 

 it underwent a permanent change with regard to the magnetic 

 chanoe of lenoth : thus it was necessarv to trv first the ex- 

 periment at low temperatures before our specimens had been 

 heated to high temperatures. 



The sensibiHty of our apparatus was such that an elonga- 

 tion or contraction of 4*8 X 10~^ of our specimen could easily 

 be observed. 



We tested six diftereut samples shown in the following 

 table : — 



Aletal. 



Length. 



Diameter. 



Demagnetizing 

 factor. 





22-05 cms. 



22-93 



22-00 



22-10 



22-03 



22-03 



1-025 cm. 



0-940 



0-948x0-953 



1044 



1-083 



1121 



0-0683 



Tungsten steel (rod)..., 

 ., (prism, 

 Ccist cobalt 



00562 

 00751 j 

 00722 i 



Ann. cobalt 



Xickel 



00758 ; 

 0-0801 ' 







§ 3. Results of Experiments. 



The change of length of our specimens at the temperature 

 of the room (8"^ C — 17^ C.) is shown in figure 3 (PI. XY.). 



