30 Joty—On the Conservation of Mass. 
above with a divided rotating head. This tube was shielded with an outer jacket of 
velvet folded several times around it. The entire apparatus was fitted up in the 
cellars of the Geological Laboratory, Trinity College, being fixed to the massive 
basement-wall of the building by a stout shelf; the temperature in this cellar 
varied slowly, seldom as much as one degree in twenty-four hours during the time 
of the experiments. The distance of the scale on which the spot of light was, 
received was such that the translatory motion of the vessels containing the active 
substances was magnified just eleven times. The divisions on the scale used are 
arbitrary; and a single division is nearly three-fifths of a millimetre. Thus a 
motion of ten divis!ons of the spot of light on the scale corresponds to a rectilinear 
displacement of the vessels amounting to alittle more than half a millimetre. The 
plane of the beam is due north and south. 
In spite of all endeavours to attain freedom from inequalities of temperature 
within the apparatus, the chief difficulty encountered in making the experiments 
has been with draughts within the case. This will readily be understood when the 
very great bulk which it is necessary to confer upon the load carried at one 
extremity of the beam is taken into account. Not only have bulky vessels to be 
there suspended, but an insulating jacket has further to be provided. Thus, in the 
experiments with which up to the present I have been principally occupied—those 
on the solution of copper sulphate in acid water—the paper cylinder within which 
the whole is contained measures 7 cms. in diameter by 8 cms. in length. Such a 
large bulk is so effectively acted upon by convection-currents within the case that 
it was, in truth, found to be never at rest, but to be continually experiencing 
slow but persistent displacements. 
Attempts to determine a consistent logarithmic decrement were not very 
satisfactory, the damping being not only rapid but occasionally irregular. The 
determination of the sensitiveness of the loaded beam to a small force applied to 
one extremity of it by a direct method was fairly successfully accomplished, as 
will presently be seen. But first the influence of the suspension as a constraint 
on a displacement from a zero position must be considered. 
The time of vibration of the fully loaded beam was found to be just thirteen 
minutes. This was with a load of 80 grammes at either end of the beam, the 
quantities of chemicals used being about one-fourth of those used in Heydweiller’s 
experiments, as will be seen later. Taking the radius of gyration of the load as 
12°5 cms., the force required to be applied to one extremity of the beam in order 
to confer a twist of one degree on the suspension calculates out as 0:00226 dynes. 
The twist is not a pure torque. The effect of this torque will be to displace the 
spot of light 2°54 ems. on the screen, or just forty divisions of the scale used. 
This corresponds to a linear displacement of the vessels through a distance of a 
little over2 mms. The effects of so feeble a restraint acting at forty divisions of 
