452 



A. Tanakadate* on the Mean Intensity of 



of the equatorial section of the bar where it will be a maxi- 

 mum, the result being independent of the mode of distribution 

 of magnetism at the broken ends. In the present case, how- 

 ever, the calculation being based upon the supposition of sole- 

 noidal distribution (which is far from being true), I might 

 be called the mean intensity of magnetization in the mass of 

 the substance under consideration. 



A glance at the diagram shows a rapid fall of mean in- 

 tensity as the length diminishes. This is no doubt due to the 

 demagnetizing effect of the broken ends, as is noticed in 

 Maxwell's book. It might be interesting here to notice how 

 this demagnetizing force reduces the areas of hysteresis as 

 the ratio length/diameter varies, although with regard to 

 this point I much regret that (except in VI.) I did not com- 

 plete the whole cycle by subjecting the bars to negative mag- 



netization. Still, the values of 



JW 



obtained from the above 



might give some notion with regard to the amount of energy 

 dissipated in carrying different bar-magnets through the 

 cycle of magnetization, 1Q being the intensity of magnetizing 

 field as usual. From considerations of results in which the 

 cycle has been completed, the areas in the above diagrams 

 will be seen to be something like one third of the whole area. 

 The areas computed from the above diagrams are given in 

 the following table and at PI. IV. fig. 7. 



Eeference 

 number. 



Length. 



Diameter. 



Length 

 Diameter' 



f70 



(•70 



f70 



3j o Id$/Jpp. 



I. 



II. 



III. 



IV. 



V. 

 VI. 



centim. 

 6 



5 



4 



3 



2 



33-4 



centim. 

 0152 



0156 



0-152 



0-152 



0-153 



0-115 



394 

 320 

 26-3 

 19-7 

 131 

 291-0 



ergs. 

 5560 



4770 



2964 



1378 



930 



ergs. 

 16680 



14310 



8892 

 4134 



2790 

 18450* 



4°-74xlO-4C. 

 4° -06 „ 

 2°-52 „ 

 1°-17 „ 

 0°-79 „ 

 5°-25 „ 



The last column of the table gives the probable rise of 

 temperature of the iron due to one complete reversal of mag- 

 netization, supposing the whole energy dissipated is used in 

 heating the iron. 



In connexion with the present result I beg to mention some 

 results of mine which were obtained in 1883 in Japan. The 



* This is the whole area of cycle J Id$. 



