Magnetism, Cohesion, Adhesion and Viscosity. 503 



two needles constructed, half of which were magnets and the 

 other half brass; the junction of the brass and steel being 

 effected exactly mid-way where the caps were attached. As 

 it was anticipated that "the movement of such a needle would 

 be slow, as the whole directive energy could only be obtained 

 from the magnetic end, their caps were of stone. I had the 

 pleasure to find that though their movement was necessarily 

 sluggish in comparison with that of the common needle, yet 

 that it was perfectly uniform and certain ; and that the brass 

 end attached itself to various substances with even more 

 tenacity than had been the case with the purely magnetic 

 needle, or with its own magnetic portion. 1 next had two 

 needles made, one with a gold and the other with a silver ter- 

 mination ; but owing to a misapprehension of the workman 

 their caps were of brass, and from this circumstance the in- 

 creased friction prevented their free movement, particularly in 

 the case of the silver needle. However, sufficient was dis- 

 covered by their employment to see that the gold and silver 

 could likewise be made to attach themselves to non-ferruginous 

 bodies. 



It was only now necessary to make such a modification of 

 the new principle of measurement as should enable it to be 

 applied to any metal, and for this purpose I had short tubes 

 soldered to the ends of half magnetic needles where they were 

 joined to their agate caps. Into these small tubes needles of 

 any malleable metal could be inserted at pleasure ; their weight 

 and length being just sufficient to balance the steel portion of 

 the needles. The steel ends were necessarily made thicker 

 than the magnetic needle I have already described, to balance 

 the increased weight of the tubes and metal needles ; and the 

 whole weighed, when complete, in one case sixteen grains, and 

 in the other, which was a little larger, nearly eighteen grains. 

 The length of the steel half was twenty-three millimetres, and 

 the whole needle therefore forty-six millimetres long. It is 

 necessary to bear in mind that the metal needles should not 

 exceed tlie magnetic ends in length, otherwise, according to 

 the principle of the lever, they would be moved with greater 

 facility than the steel portion, and give proportionably higher 

 results. 



The Ibrm of these needles is here given with and without 

 their metal ends. 



By a further modification, the arrangement was fitted to be 



