64 
TRANSACTIONS OF THE TEXAS ACADEMY OF SCIENCE. 
of heavy wire, terminating in knobs. Lines marked 1, 2, 3, are drawn 
through the nodal points as respectively found by these three resonators. 
It will he observed that the nodal' distances found by ISTo. 3 are most dis- 
tant, and those found by No. 1 least distant. With smaller resonators 
these differences are proportionately greater, and this use of knobs is 
quite sufficient to explain the error into which Sarasin and De La Rive 
fell when they concluded that these distances are always four times the 
diameter of the circle. If their law had held true in this particular case, 
these distances would have been 633 c.m. The greatest distance given 
by No. 3 was 590, with an average of 540, while with No. 1 it varied from 
480 to 533.* 
In these experiments the resonator was held vertically, its plane in- 
cluding the wire. This position was chosen in preference to the one at 
right angles by reason of the fact that it removed the resonator from the 
direct influence of the oscillator and was the one deemed least liable to 
the effects of reflections. One end of the wire under examination was 
secured to a plate 16 c.m. square, suspended opposite and quite near one 
of the plates of the vibrator, and the other end was held in place by a 
silk cord. Beginning near the free end of the wire, the spark gap of the 
resonator was drawn out to near the maximum sparking length, and the 
resonator was then carried forward along the wire. Sometimes it was 
soon extinguished; again could be carried the whole length without any 
appreciable change. In the former event the gap was made smaller, and 
in the latter greater. After a few trials a particular spark could be ob- 
tained that would continue up to within a short distance of a certain point 
as it was approached from either side. The point of disappearance was 
recorded by a chalk mark upon the floor, the room being so dark that the 
record could not be seen. After quite a number of such observations had 
been made they were examined, and if found well grouped together the 
average distance to the end of the wire was measured; otherwise, they 
were rubbed out and the observations repeated. Sometimes the extreme 
records would be only an inch or two apart; again they would be scattered 
over a yard or more, and despite the greatest pains and repeated efforts 
would remain so. For the sake of explaining the diagram, we will sup- 
pose that the experiments begin with the wire 24 meters in length, the 
oscillator being at the end terminating on the line marked AB. With 
resonator No. 1, made of 24 wire, and having free ends, four nodes are 
*It should have been stated that the data for these resonance curves were 
obtained in several different rooms, of different sizes, and on different floors; 
most of the work having been done in a third-floor room 90x130 feet. In all cases 
the results were the same. It is therefore believed that reflections have not 
played any important part in the result. 
