On page 3, Russell concludes his analysis of his experience of the table: "The real
table, if there is one, is not immediately known to us at all, but must be an inference
from what is immediately known. What main consideration leads Russell to this
conclusion?