90 
Proceedings of the Royal Society 
By keeping right contact for decreasing rotation- swing, and 
left for increasing, the range of swing increased after 6 swings to 
123 divisions. On reversing this action for 6 swings the range 
was reduced to 12 divisions. 
Note. — These are the most striking experiments. It will be 
noted that here the “ right ” gives lower readings than the “ left ” 
contact. But I cannot be sure that the commutator arrangements 
had not been changed. Here all disturbing effects are done away 
with, and the right and left contacts produce deflections on 
opposite sides of the “ No contact ” reading. 
No contact, 
Right „ 
Left „ 
No 
5 ) 
Flint-Glass in Water. 
„ . 541 scale reading. Means. 
531 R = 531. 
551 N = 541. 
540 L = 551. 
A new rod was now prepared, which also showed a very decided 
difference between right and left contact, right being lowest. 
Numerous experiments under various conditions were now tried, 
but no new effects were discovered. It is desirable, however, to 
record the following, which were made to be certain as to the 
direction of rotation under different conditions of polarity : — 
No contact, ....... 
850 
Upper pole would point to north if free, 
840 
No contact, ....... 
848 
Lower pole, 
854 
Upper pole, 
835 
Lower pole, 
850 
No contact, ....... 
845 
The general result of the experiments of the winter is, that 
when a 'pole tending to point to the north is above the glass rod , and a 
pole tending to point to the south is below , the rod turns in the opposite 
direction to the hands of a watch with the face upwards. 
On placing the magnet with the poles in a horizontal plane, 
rotations were sometimes observed. This is in direct contradiction 
to the theory that led to these experiments. I can offer no 
explanation, but merely record the fact. 
In the session 1873-74 similar experiments were made with a 
