37G Dr. Latnont on the most advantageous Form of Magnets. 

 ments is calculated by the formula 



V 



12-80 + 2--JL6 n 

 n +'0-218 n ' 



Number of 



Magnetic 



moment. 



laminse=n. 



Observed. 



Calculated, 



1 . . 



. 3-53 



351. 



2 . . 



. 4-11 



4-00 



3 . . 



. 4-36 



434 



4. . 



. 465 



4-63 



5 . . 



. 4-9 1 



4-90 



6 . . 



. 515 



5-1G 



7. . 



. 5-39 



5-40 



8. . 



. 5-61 



562 



9. . 



. 5-83 



5-84 



10. . 



. 605 



6-05 



11 . . 



. 6-27 



6-25 



12. . 



. 0-41 



6-45 



Difference. 



+ 001 

 ^011 

 -0-02 

 -0-02 

 -004 

 + 001 

 + 001 

 + 001 

 + 0-01 

 000 

 -002 

 + 001 



A practical inference results from the preceding investigation, 

 which I believe deserves to be carefully considered on the part 

 of those who concern themselves with the manufacture of mag- 

 netic instruments. A freely moveable magnet is to be employed 

 with advantage only so far as the magnetic moment is as large 

 as possible in proportion to the weight. But the more the trans- 

 verse size is augmented, the greater is the departure from the 

 fulfilment of this condition, and consequently the use of massive 

 magnetic bars must be pronounced inadmissible. There is only 

 one means of obtaining great magnetic strength with trifling 

 weight ; namely, by firmly connecting several thin and flat mag- 

 nets near or upon one another in one system without their 

 touching each other. Many years ago I began in magnetic 

 variation-instruments, later also in magnetic theodolites, to unite 

 several magnets ; and at present I use universally systems of 

 three laminae; which are laid upon each other and held separated 

 in the middle by small pieces of brass of about the thickness of 

 three-quarters of a line. Also in ships' compasses several needles 

 near each other ar^ at present continually used with the best 

 result. Hollow cylindrical magnets, to which some artists have 

 given a great preference in regard to strength and lightness, 

 remain, as can be proved even from theoretical considerations, 

 very far behind in comparison with a single flat needle ; and 



