158 Intelligence and Miscellaneous Articles. 



piece of apparatus, which may be called a computer, consisting of 

 a sliding-rule, index-arm, and scale. Suppose we wish to divide 

 accurately the space x 19 formerly given ; we set the distance from 

 P to Q on the slide, to that of Q to E on the fixed scale, and the 

 index-arm at once points to the angle on the scale which M N must 

 make with A C on our divided triangle. 



In use, this computer is found both trustworthy and simple, 

 and facilitates w T hat would otherwise have been tedious to perform 

 and liable to inaccuracies when done. 



Having devised a method of dividing a thermometer-tube accu- 

 rately, I next turned my attention to the most suitable strength 

 of alcohol to be employed in filling these thermometers ; and as 

 but little can be gained by referring to our scientific writers on 

 this subject, I undertook a series of experiments to satisfy myself 

 on this point. 



Now it is well known that the expansion of alcohol by heat is 

 not proportionate to its change of temperature ; and we are con- 

 sequently unable to express its corresponding coefficient by a 

 simple formula, as for metals and mercury, but must adopt the 

 more complex expression of M. Biot, 



d=at-\-bt 2 , 

 where a and b are the required coefficients, and t the temperature. 

 The coefficients of absolute alcohol are constant for all tempera- 

 tures, but inasmuch as this fluid becomes weaker from momen- 

 tary exposure to the air, and the mere pouring from one bottle to 

 another not only alters its specific gravity but also its rate of 

 expansion and its coefficient, it becomes an important point to 

 select some other strength less liable to those changes, which only 

 tend to inaccuracies, however much we may strive to prevent 

 them. In making this selection, various specific gravities and of 

 varying purities were tested, and the results led to the selection 

 of alcohol having a specific gravity of '815, as spirit of this strength, 

 if pure, and free from methylated spirit, fusel-oil &c, has its co- 

 efficients as constant as those of absolute alcohol, and it is but 

 little liable to undergo those changes that seem to thwart our 

 efforts when using a stronger sample. — From the Proceedings of 

 the Meteorological Society for November 20, 1867. 



DIALYSIS OE INDUCED CURRENTS. 

 NOTE BY M. E. BOUCHOTTE. 



The machine for induction-currents which was used in the fol- 

 lowing experiments is from the works of the Alliance Company. 

 It has eight bobbins, each with a coil of 160 metres of copper wire 

 1 millim. in diameter. Eight horseshoe magnets act upon the bobbins, 

 and at each turn of the axis produce eight positive and as many ne- 

 gative currents. It is worked day and night by an hydraulic wheel ; 

 it can be moved at rates between 250 and 800 turns in a minute ; it 

 has hitherto been kept at a velocity of 500 turns. 



We first endeavoured to determine the electromotive force of the 

 apparatus. For this purpose it was provided with a commutator 

 which turned all the currents in the same direction. The circuit 



