With the sense of hearing exercised at its best, I continued the search, but had almost given it up when, upon crossing the room and passing under the chandelier, I thought I heard the jangle above my head. Getting upon a chair, I listened carefully at each one of the glass globes, and finally came to one where I could hear the jangle quite distinctly. Upon looking at this globe carefully I discovered a very peculiar crack in it. This crack in shape was almost a complete circle, but a small stem or portion of glass at its lower edge held the piece in place, so that it was in condition to respond to the vibrations of that note of the piano with which it was in tune, and in this case it was the one that was being sounded.
The accompanying sketch will give a very fair idea of the position and shape of this piece of glass at (A). Pressing a finger against it in order to stop its vibrating, and to be quite sure that 
I had found the trouble, I asked the player if she heard the sound then. After several vigorous thumps she was obliged to confess that she did not. Taking away my finger and allowing it to vibrate as before, I asked again, "Do you hear it now?" The answer this time was, "Yes, it is there yet." Removing the globe, I announced the fact that the piano was fixed, much to the astonishment of the player, who found the statement correct. This incident illustrates how even the practiced ear of a musician can sometimes be deceived as to the location of sounds in music—to say nothing of the ladies, who would be excusable under such circumstances, as their sense of hearing is not expected to be so perfect as to detect such peculiar phases of sound.
Another interesting feature of peculiar sound effects is illustrated by this incident. While this loose piece of glass would respond and vibrate to one note of the piano, no other note would affect it, not even the sharp or flat of the one that caused it to respond.
If these responding objects, however, were free to vibrate without touching anything, such as a violin string, there would be no jangle, for as in the above case the edges of the broken piece of glass touched that of the globe, which caused the discordant sound, which I have termed for want of a better name the responsive jangle.
Another case similar to the above occurred in a house where I was once stopping in Nova Scotia. A piano with a bad note was fixed by simply opening an inside shutter of a bay window at the opposite side of a parlor from the piano. The latch of one shutter was lightly resting against the edge of another and
