417 
1910-11.] A Modified Form of Atwood’s Machine. 
The results here appended are not to be taken as giving the limits of 
accuracy of the method, for, unfortunately, the construction of the friction 
rollers in the apparatus used is not quite satisfactory, with the result that 
friction somewhat variable and in need of constant evaluation : with a 
more carefully constructed apparatus friction would he much smaller, and its 
constancy more assured. 
2. Description of Apparatus. (See fig.) 
A six-spoked aluminium wheel, cut with a Y-groove in the rim, and 
mounted on a steel spindle, runs on four aluminium friction rollers. These 
rollers have conical sockets and are mounted on points screwed into a 
brass plate, so that if the steel spindle were completely homogeneous and 
were laid on the rollers, electrical contact would be complete from one 
side of the apparatus to the other during one whole revolution of the 
wheel. At one end of the spindle, however, a semi-cylindrical portion of 
the steel is removed and replaced by an identical piece of hard ebonite, 
the whole being then turned true in the lathe ; it follows that at a certain 
point during each revolution the ebonite is in contact with both friction 
rollers on which it rests, and at that moment electrical contact is no longer 
possible between the two sides of the apparatus. Hence, by connecting 
with a chronograph, a single record is obtained of the distance fallen 
through during each revolution of the wheel (being the distance equal 
to the effective circumference of the wheel and string), and of the time 
taken to describe that distance. A sensitive relay is interposed between 
the revolving wheel and the chronograph, so that a very small current may 
be used, and no sparking occurs during motion. In the experiments given 
in this paper a three-pen motor-driven chronograph was used in conjunction 
with a clock beating half seconds, and times were easily obtained to T -J(yth 
of a second. The third pen served to record the actual moment at which 
motion began. Its -magnet is connected in parallel with a small electro- 
magnet held in an adjustable stand. When an experiment is about to be 
made, the pan with the smaller load, to the bottom of which is soldered 
a small iron disc, is brought down to this release magnet, and, if necessary, 
its height is adjusted until the point of no contact in the wheel revolution 
is just reached. Paper is inserted between the magnet and the pan to 
ensure immediate release ; the current is then broken, setting the system 
in motion and recording the initial point of the experiment on pen No. 3 ; 
every complete revolution of the wheel is recorded on pen No. 2, while the 
half seconds are marked by pen No. 1. 
VOL. XXXI. 
27 
