314 



Scienfific Proceedings^ Eoijal Dublin Societij. 



From the curve in fig. 1 we see that as the dimension ratio increases the 

 magnetic moment per gramme increases up to a certain point, and then tlie 

 curve becomes flat, showing that there is one value of the dimension ratio, 

 or one value of the length of magnet for a given diameter tiiat gives the 

 best result. That is, in tliis curve the dimension ratio 80, or a magnet of 

 length 12'8 cms., is the best length to make witli this steel of 0-16 em. 

 diameter. 



The same operations were gone through for other eleven samples of the 

 same kind of steel, i.e. six magnets were made of eacli, tempei'ed, cleaned, 

 weighed, and tested for magnetic moment per gramme. 



A curve for each set was drawn like fig. 1, and the most effective value 

 of the dimension ratio taken from the curve, and the best length of magnet 

 ascertained. 



Tlie results are shown in Table II. ; and if we were to plot the values in 

 column 2 of the Table as abscissae, and tlie corresponding values in column 3 

 as ordinates, the curve obtained would show that from a diameter of 

 0"16 cm. to about 0'5 cm. the points lie very approximately in a straight 

 line, whereas for diameters larger than 0'5 cm. or up to 1 cm. the curve 

 droops sliglitl}' towards the origin. 



