TRANSACTIONS Of SECTION A. 



583 



The scale is calibrated for a standard size of specimen, and the value for any- 

 other size is obtained by multiplying by the ratio. A wide range is obtained by 

 using only a part of the magnetising coils when the permeability is high. 



The magnetising force is read on a separate instrument, such as a suitable 

 amperemeter. Regulating resistances, a reversing switch for demagnetising, and 



a switch for altering the range are added. 



11. Note on the Coherer, By Professor James Blyth, M.A.^ LL.D. 



The object of this note is to draw attention to some experimental results con^ 

 nected with the ordinary filings-coherer, which I caa hardly think are new, but 

 which I have not seen specially noticed. 



When a coherer is placed in circuit with a battery, and when no current 

 passes through it, it is obvious that its terminals must coiTespond to the charged 

 plates of a condenser, and that the P.D. between them 

 must be equal to the E.M.F. of the battery. Let AB 

 be the coherer and C the battery, then the P.D. between 

 A and B is equal to the E.M.F. of 0. If now A and B be 

 connected for an iustani by a circuit containing a coil E 

 having self-induction, the coherer AB will be found to 

 have assumed the conducting instead of the insulating 

 condition. This can be tested by switching a galvano- 

 meter into the battery circuit and observing the deflec- 

 tion. If, however, the coherer AB be short-circuited 

 for an instant by a coil having the same resistance as E, 

 but wound so as to have no self-induction, the coherer 

 does not become a conductor. 



This would seem to show that the discharge of the 

 condenser-coherer must be of a distinctly oscillatory nature before the well-kuown 

 effect of coherence is produced. 



The next result I have to refer to depends essentially on the same cause. 



Let two coherers AB and CD be included in the same circuit with a 

 battery F and a galvanometer or bell G. Also let a Voss machine be placed 

 near AB so as to produce an 

 oscillatory spark near AB, but 

 let CD be placed so far away 

 as to be beyond the direct 

 action of the spark ; then in 

 general it will be found that 

 when AB becomes a conductor 

 suddenly the jerk given to CD 

 is sufficient to make it also a 

 conductor, and the galvanometer will deflect or tlie bell ring. If CD be now 

 tapped, the bell stops, although AB has been left untouched. This shows that 

 if one coherer in a circuit suddenly assumes the conducting condition all other 

 coherers in the same circuit tend to do the same. 



Q q2 



