298 The Rev. T. R. Robinson's Experimental Researches on the 



during a series of measures. The least of the sliding weights is now hung to 

 its loop and cautiously moved, till either it lifts the keeper, or arrives at the 

 end of the division. In the latter case it is changed for a heavier. If none of 

 them overcome the magnet, a scale weight, equivalent to the greatest moment 

 of the last of the steelyard weights, is placed in the dish, and so on. Those 

 which I use are Ol lb., 0-2, and 0-6 for the steelj'ard ; the others are 0'5, 1, 2, 

 4, 7, and 14 ; the dish also = 0-5. They were carefully verified by a set of 

 grain-weights belonging to me, and another of Professor Stevellt. The le- 

 verage of the machine was determined with equal care. By means of the above 

 weight and a balance, for the use of which I am indebted to my friend Mr. 

 Mallet,* two of 28 lbs. and two of 56 were verified. Suspending them to 

 the keeper, I found the weights required to counterpoise 56, 112, 168, and 199 

 lbs., and obtained their ratio — 59-730. In these trials additions to the load of 

 0-031, 0-046, and 00941b. were easily detected; an error of about lib. in 

 the ton. 



A machine of this kind is of course not expected to equal the accuracy uf 

 an ordinary balance ; but for the work which it has to do it is far preferable 

 on two accounts. To lift the keeper by weights equal to its attraction would 

 be very dangerous, for the sudden descent of 8 or 9 cwt. would cause a 

 fearful concussion ; while the fall of its equivalent, 15 on the pad T', is scarcely 

 felt. Besides, when the separation is nearly attained, the most delicate mani- 

 pulation is necessary ; and it is far easier to avoid jar in sliding a light weight, 

 than in placing in a scale one sixty times as heavy. But in fact the force to be 

 measured is itself fluctuating to an extent which far passes any errors of the 

 weighing. 



3. I have measured the intensity of the voltaic current by a tangent 

 rheometer; and this mention of it niiglit suffice, were it not that even in an in- 

 strument so well known the details of its use are not without value, and that its 

 results cannot be duly compared to those of another without a distinct know- 

 ledge of its individuality. I prefer it to the one described in a former commu- 



* It 13 the smallest of those mentioned in his Report to the British Association on the Corro- 

 sion of Railway Bars; when loaded with 56 lbs. in each scale it turns decidedly with three grains. 

 All these comparisons were made by the method of double weighing. 



